VIRTOPSY APPROACH
The
3D Optical and Radiological Scanning and Reconstruction in Forensic Medicine
© 2009 by Taylor ...
57 downloads
573 Views
31MB Size
Report
This content was uploaded by our users and we assume good faith they have the permission to share this book. If you own the copyright to this book and it is wrongfully on our website, we offer a simple DMCA procedure to remove your content from our site. Start by pressing the button below!
Report copyright / DMCA form
VIRTOPSY APPROACH
The
3D Optical and Radiological Scanning and Reconstruction in Forensic Medicine
© 2009 by Taylor & Francis Group, LLC
VIRTOPSY APPROACH
The
3D Optical and Radiological Scanning and Reconstruction in Forensic Medicine Edited by
Michael J. Thali, M.D. Richard Dirnhofer, M.D. Peter Vock, M.D.
Boca Raton London New York
CRC Press is an imprint of the Taylor & Francis Group, an informa business
© 2009 by Taylor & Francis Group, LLC
CRC Press Taylor & Francis Group 6000 Broken Sound Parkway NW, Suite 300 Boca Raton, FL 33487-2742 © 2009 by Taylor & Francis Group, LLC CRC Press is an imprint of Taylor & Francis Group, an Informa business No claim to original U.S. Government works Printed in the United States of America on acid-free paper 10 9 8 7 6 5 4 3 2 1 International Standard Book Number-13: 978-0-8493-8178-2 (Hardcover) This book contains information obtained from authentic and highly regarded sources. Reasonable efforts have been made to publish reliable data and information, but the author and publisher cannot assume responsibility for the validity of all materials or the consequences of their use. The authors and publishers have attempted to trace the copyright holders of all material reproduced in this publication and apologize to copyright holders if permission to publish in this form has not been obtained. If any copyright material has not been acknowledged please write and let us know so we may rectify in any future reprint. Except as permitted under U.S. Copyright Law, no part of this book may be reprinted, reproduced, transmitted, or utilized in any form by any electronic, mechanical, or other means, now known or hereafter invented, including photocopying, microfilming, and recording, or in any information storage or retrieval system, without written permission from the publishers. For permission to photocopy or use material electronically from this work, please access www.copyright.com (http://www.copyright.com/) or contact the Copyright Clearance Center, Inc. (CCC), 222 Rosewood Drive, Danvers, MA 01923, 978-750-8400. CCC is a not-for-profit organization that provides licenses and registration for a variety of users. For organizations that have been granted a photocopy license by the CCC, a separate system of payment has been arranged. Trademark Notice: Product or corporate names may be trademarks or registered trademarks, and are used only for identification and explanation without intent to infringe. Library of Congress Cataloging-in-Publication Data The virtopsy approach : 3D optical and radiological scanning and reconstruction in forensic medicine / editors, Michael J. Thali, Richard Dirnhofer, Peter Vock. p. ; cm. Includes bibliographical references and index. ISBN 978-0-8493-8178-2 (hardcover : alk. paper) 1. Forensic radiography. I. Thali, Michael J. II. Dirnhofer, Richard. III. Vock, Peter. IV. Title. [DNLM: 1. Autopsy--methods. 2. Forensic Medicine--methods. 3. Imaging, Three-Dimensional--methods. 4. Radiographic Image Interpretation, Computer-Assisted--methods. W 825 V819 2008] RA1058.5.V57 2008 614’.1--dc22 Visit the Taylor & Francis Web site at http://www.taylorandfrancis.com and the CRC Press Web site at http://www.crcpress.com
© 2009 by Taylor & Francis Group, LLC
2008013717
Dedication Only ideas and committment are changing the scientific world
© 2009 by Taylor & Francis Group, LLC
Table of Contents Part A
Introduction ........................................................................................................................................................... 1
Chapter A1 From Autopsy to Virtopsy: Oral Description versus Image: Value of Evidence .................................................. 3 Chapter A2 History of Virtopsy: How It All Began .............................................................................................................. 11 Chapter A3 Virtopsy® after More Than Some 100 Cases: Statement and Perspectives of Forensic Imaging by Using 3D Optical and Combined CT/MRI Whole-Body Scanning .............................................................. 19 Chapter A4
Legal Side ........................................................................................................................................................... 27
Chapter A5 Religion ............................................................................................................................................................... 41 Part B
Imaging and Visualization Methods/Explanation of Techniques....................................................................... 49
Chapter B1
External Body Documentation ........................................................................................................................... 51
Chapter B2
Internal Body Documentation ............................................................................................................................ 61
Chapter B3
3D Visualization of Radiological Data ..............................................................................................................115
Chapter B4
Storage of Radiological Data (PACS) ................................................................................................................131
Chapter B5
The Virtopsy Database: Comparing Radiology and Autopsy Findings Using a Database ............................. 135
Part C
Forensic Application of Imaging Techniques ....................................................................................................143
Chapter C1 Intravital versus Postmortem Imaging .............................................................................................................. 145 Chapter C2 A Historical Overview of the Literature ............................................................................................................147 Chapter C3 External Body Documentation ......................................................................................................................... 151 Chapter C4 Internal Body Documentation .......................................................................................................................... 157 Chapter C5 Documentation of Extracorporeal Findings ..................................................................................................... 159
© 2009 by Taylor & Francis Group, LLC
Part D
Forensic Topics ................................................................................................................................................. 167
Chapter D1
Radiologic Identification .................................................................................................................................. 169
Chapter D2 Thanatology ...................................................................................................................................................... 187 Chapter D3 Incident-Specific Cases......................................................................................................................................219 Chapter D4 Virtopsy as a Multi-Tool Approach .................................................................................................................. 389 Chapter D5 Biopsy .............................................................................................................................................................. 437 Chapter D6 Postmortem Angiography ................................................................................................................................. 443 Chapter D7 Experiences with Virtual Autopsy Approach Worldwide ................................................................................ 475 Chapter D8 Miscellaneous ................................................................................................................................................... 479 Acknowledgments ................................................................................................................................................................... 501
© 2009 by Taylor & Francis Group, LLC
Editors Michael Thali, M.D., Executive MBA HSG, has worked since 1995 in the field of forensic medicine. He had a two-year fellowship in clinical radiology. In 2001/ 2002 he was a fellow at the office of the Armed Forces Medical Examiners at the Armed Forces Institute of Pathology (AFIP) in Washington, D.C. He has written many virtual autopsy papers (see www.virtopsy.com). Since February 2006 he has been full professor of forensic medicine at the University of Bern, Switzerland. He is director of the Forensic Institute and the Center for Forensic Imaging at the University of Bern. Richard Dirnhofer, M.D. has worked since 1967 in the fields of pathology and forensic medicine. Since 1974 he has been deputy director of the Institute of Forensic Medicine in St. Gallen, Switzerland, and since 1979 at the Institute of Forensic Medicine at the University of Graz. From 1984 to 2005 he was full professor for forensic medicine, first at the University of Basel and then Bern, and was visiting professor at the University of Salzburg. His main foci
© 2009 by Taylor & Francis Group, LLC
in forensic research are forensic pathology, forensic DNA analyses, and in more recent years the field of virtopsy. He was president of the Swiss Society of Legal Medicine and 1999 initiator of the Virtopsy Project. In 2003 he founded the Virtopsy Foundation and, together with Professor Thali and Professor Vock, founded the Technical Working Group Forensic Imaging Methods (TWGFIM). Since 2005 he has been professor emeritus of legal medicine. Peter Vock, M.D. is Professor of Radiology and Chairman, Institute of Diagnostic, Interventional and Paediatric Radiology, Inselspital, University of Bern, Switzerland. He has worked since 1974 in radiology, with fellowships in nuclear medicine and radiotherapy, and was a visiting research associate at Duke University Medical Center. Together with Willi Kalender, he co-invented spiral computer technology. He is a board member of many international radiological societies. He has made his Radiology Institute available to the Virtopsy project and has intensively supported the idea from its inception.
Preface (R)EVOLUTION IN FORENSIC MEDICINE Forensic medicine is an important endeavor with great significance in public society. Ultimately, the forensic standard defines the status quo of the rule of law. In his 1998 book Forensic Radiology, Gil Brogdon wrote that there is no definition or standard for forensic radiology. There is also no specialization or training fellowship in this field; there is no Society for Forensic Radiology and no subdiscipline. Similar to how the forensic disciplines of chemistry, toxicology, and molecular biology have risen in the last century in the traditional fields of clinical forensic medicine and postmortem forensic medicine, a change process has begun in that, with modern imaging procedures, added value and quality improvement could be engendered. In the past few years the transdisciplinary international Virtopsy Team
© 2009 by Taylor & Francis Group, LLC
of the University of Bern has brought about numerous publications in this area. We have been approached from many sides to summarize our knowledge. The first attempt in this direction has resulted this book. Statements about the future are difficult to make, but it is indisputable that imaging procedures will take on a rapidly increasing importance within all of forensic medicine. Whether with the so-called virtopsy approach is an evolution or a revolution in forensic science, the reader will ultimately have to decide for him- or herself, based on his or her observations of the coming developments in this field. Michael Thali Richard Dirnhofer Peter Vock
Foreword With this book the “Virtopsy” group illustrates and explores the expanding dimensions of a great new era in the application of medical imaging in the forensic sciences. Although fresh, it is long overdue. In 50 days of intensive investigation after first observing the x-ray on November 8, 1895, Röntgen defined the basic properties of this neue Arte von Strahlen so thoroughly that decades passed before anything new could be added to the subject. Presentation and publication of his findings were delayed by the Christmas holidays until January 23, 1896, but a little of the scientific “leakage” so familiar to us today led to disclosure of Röntgen’s discovery in the worldwide press on January 5 and 6 of that new year. Some of those accounts were lurid and inaccurate. Yet many physicians, scientists, jurists, and responsible journalists understood the great potential of Röntgen’s Ray to resolve a wide variety of forensic issues. A majority of modern usages of radiology in the forensic sciences were predicted or practiced by 1898: adulterations of foodstuffs, archeological/anthropological investigations, attempted murder, bomb detection, bone age determination, cause and manner of death, celebrity roentgenography, dental identification, fingerprints, forgery, fraud, larceny by ingestion, liability, medical malpractice, missile identification/ localization/extraction, murder, nondestructive testing/industrial radiography, personal injury, postal fraud, radiological malpractice, skeletal identification, and smuggling. Some of these early applications became obsolete; others fell into disuse or were ignored for years; some would have a resurgence years later (e.g., skeletal identification). The only really new applications of medical radiology in the forensic sciences in half a century were in mass casualty situations and in the recognition of physical abuse in children, the elderly, and intimate partners. So for a hundred years forensic radiology was essentially a stagnant field. While radiologic apparatuses improved, there was little need for them to study material that neither moved, breathed, peristalsed, nor pulsated. Radiological knowledge improved, but few radiologists were interested in forensic work, and still fewer actually did any. Most forensic radiology was carried out by pathologists, dentists, or anthropologists on antiquated equipment assisted by untrained ancillary personnel. (Abraham Lincoln’s head on the penny seemed to be the most widely used standard of film quality and exposure technique in the United States. I don’t know what coin may have served that purpose in other countries.) Forensic images were mostly in case reports scattered in the journals of many disciplines. There was no common fund of radiologic knowledge and no comprehensive body of literature available to those practioners who needed radiologic capability for their individual forensic activities.
© 2009 by Taylor & Francis Group, LLC
Exciting new developments in the radiological armamentarium came about in the last quarter of the twentieth century, notably ultrasonography (US), computed tomography (CT), and magnetic resonance imaging (MRI). The latter two sectional imaging modalities looked especially appealing for forensic investigation. But the equipment was expensive, required skilled operators and maintenance, and was almost totally occupied in the provision of clinical medical services. Few forensic scientists found these tools available, and later (largely unfounded) fears of contaminated body fluids virtually terminated off-hours use of hospital facilities for postmortem investigations. Only a few forensic applications of the new modalities were published, mostly in case studies driven by necessity when only sophisticated antemortem images were available for comparative identification of remains. A very few limited prospective studies were undertaken in institutions where accessibility and availability could be afforded on a limited-time basis. Thus in 1998, I was to write, “The sad truth is that a century after the first x-ray was introduced in a court of law, there is no general appreciation of the extent of the radiologic potential in the forensic sciences.” This plaintive note apparently struck a chord in the soul and mind of Richard Dirnhofer, M.D., then professor and director of the Institute of Forensic Medicine at the University of Bern. He and Peter Volk, M.D., professor and director of the Institute of Diagnostic Radiology in that same medical center, established a uniquely organized collaborative interdisplinary group from their two institutes to undertake the definitive study of the utility of sophisticated imaging methods in the investigation of death. This project, a model for interdepartmental and interdisciplinary cooperation, has been enormously successful in producing a prodigious body of scholarly publications from which this volume is drawn. They have attracted additional collaborators and contributors from other institutions and countries, and have stimulated intercontinental interests in similar investigations. Professor Dirnhofer created a neologism to identify the project and the process of minimally invasive, imagingguided virtual autopsy involving several modalities and intricate postprocessing by combining “virtual” and “autopsy” into the conveniently short but descriptive word, virtopsy. (No small achievement for a man whose mother tongue is German, in which most new words seem to be created by combining an endless stream of old words!) The Virtopsy Approach: 3D Optical and Radiological Scanning and Reconstruction in Forensic Medicine and Science is an up-to-the-minute, commodious summary of the Virtopsy experience, well organized, detailed enough to serve as a how-to guide for newcomers to the field, copiously illustrated with many color figures accompanied by appropriate
explanatory captions. The contributors furnish state-of-theart expertise in their individual specialties. The spectrum of technologies employed is comprehensive: ultrasonography, computed tomography, magnetic resonance imaging and spectroscopy, photogrammetry, and surface scanning. The scope of the Virtopsy Project, from conventional autopsy to veterinary science, will surprise most readers, and there are prophetic glimpses of new horizons just now coming into view. The virtopsy group has been careful to characterize their work as supplemental to the traditional open autopsy, stressing the advantages and failings of both procedures. But there are situations when virtopsy can and will substitute for conventional autopsy where law, religion, or cultural mores prohibit invasion of the body. Already virtopsy is attracting great interest in some of those jurisdictions and societies. Another important area in which virtopsy could substitute quite valuably for conventional autopsy is in education. Firstly, virtopsy could be used to teach gross anatomy where cadaveric material is scarce or unavailable. Secondly, for generations, the autopsy rate has been a matter of prime interest, importance, and pride in medical schools and teaching hospitals. The classic Clinico-Pathology Conference (the time-honored CPC) was a weekly, widely-attended dramatic educational showpiece in those institutions. The contemporary non-forensic autopsy rate in teaching hospitals, for a variety of reasons, ranges from abysmal to zero. A virtopsy-
© 2009 by Taylor & Francis Group, LLC
guided CPC could be an intellectually and visually exciting replacement. Finally, in several countries or on several continents, sometimes spontaneously, sometimes perhaps influenced by the Bern reportage, postmortem sectional imaging is ongoing. Reports of such activities are coming from Germany, France, the UK, Scandinavia, Japan, Australia, and Israel. The United States has lagged behind in this, but intensive multimodality investigative protocols now under way at the Armed Forces Mortuary when published may spur belated efforts in the private sector to catch up with the other privileged nations of our world. Professor Thali, protégé of and successor to Professor Dirnhofer, is to be congratulated for his gifted editing and leadership in the production of this book. The offerings of his contributing authors are equally commendable. It has been my privilege to follow the activities and success of the Virtopsy Project almost from its inception. Consequently, I am most honored and pleased to furnish the foreword of this truly remarkable and seminal volume. Gil (B.G.) Brogdon, M.D. University Distinguished Professor Emeritus of Radiology The University of South Alabama College of Medicine Mobile, Alabama
Contributors Emin Aghayev Institute of Forensic Medicine Centre for Forensic Imaging and Virtopsy University of Bern Bern, Switzerland Lea Attias Institute of Forensic Medicine Centre for Forensic Imaging and Virtopsy University of Bern Bern, Switzerland Marcel Braun Institute of Forensic Medicine Centre for Forensic Imaging and Virtopsy University of Bern Bern, Switzerland Chris Boesch Department of Clinical Research MR Center 1 University and Inselspital Bern, Switzerland Stephan A. Bolliger Institute of Forensic Medicine Centre for Forensic Imaging and Virtopsy University of Bern Bern, Switzerland B.G. Brodgon College of Medicine University of South Alabama Mobile, Alabama Ursula Buck Institute of Forensic Medicine Centre for Forensic Imaging and Virtopsy University of Bern Bern, Switzerland
Silke Grabherr Institute of Forensic Medicine University of Lausanne Lausanne, Switzerland Michael Ith Institute of Forensic Medicine Centre for Forensic Imaging and Virtopsy University of Bern Bern, Switzerland Christian Jackowski Institute of Forensic Medicine Centre for Forensic Imaging and Virtopsy University of Bern Bern, Switzerland Willy A. Kalender Institute of Medical Physics University Erlangen-Nürnberg Erlangen, Germany Marek Karolczak Institute of Medical Physics University Erlangen-Nürnberg Erlangen, Germany Beat P. Kneubuehl Institute of Forensic Medicine Centre for Forensic Physics and Ballistics University of Bern Bern, Switzerland Patric Ljung Siemens Corporate Research Princeton, New Jersey Claes LundstrÖm Center for Medical Image Science and Visualization Linköping University Sweden
Andreas Christe Institute of Forensic Medicine Centre for Forensic Imaging and Virtopsy University of Bern Bern, Switzerland
Silvio Näther Institute of Forensic Medicine Centre for Forensic Imaging and Virtopsy University of Bern Bern, Switzerland
Richard Dirnhofer Institute of Forensic Medicine Centre for Forensic Imaging and Virtopsy University of Bern Bern, Switzerland
Lars Oesterhelweg Institute of Forensic Medicine Centre for Forensic Imaging and Virtopsy University of Bern Bern, Switzerland
© 2009 by Taylor & Francis Group, LLC
Anders Persson Center for Medical Image Science and Visualization Linkoping University Linkoping, Sweden Kimberlee Potter Office of the Armed Forces Medical Examiner Armed Forces Institute of Pathology Rockville, Maryland Steffen Ross Institute of Forensic Medicine Centre for Forensic Imaging and Virtopsy University of Bern Bern, Switzerland Eva Scheurer Institute of Forensic Medicine University of Graz Graz, Austria Graham P. Segal Sydney, Australia Danny Spendlove Institute of Forensic Medicine Centre for Forensic Imaging and Virtopsy University of Bern Bern, Switzerland
© 2009 by Taylor & Francis Group, LLC
Michael J. Thali Institute of Forensic Medicine Centre for Forensic Imaging and Virtopsy University of Bern Bern, Switzerland Peter Vock Inselspital Department for Diagnostic Radiology University of Bern Bern, Switzerland Markus Weber Fürsprecher Generalprokurator Kanton Bern Bern, Switzerland Kathrin Yen Institute of Forensic Medicine University of Graz Graz, Austria Anders Ynnerman Division for Visual Information Technology and Applications Linkoping University Linkoping, Sweden
Part A Introduction
© 2009 by Taylor & Francis Group, LLC
A1
From Autopsy to Virtopsy: Oral Description versus Image: Value of Evidence Richard Dirnhofer
CONTENTS A1.1 About Preparation .............................................................................................................................................................. 5 A1.2 About Determination of the Findings ................................................................................................................................ 5 A1.3 About Documenting the Findings ..................................................................................................................................... 6 A1.4 Concerning the Term Virtopsy .......................................................................................................................................... 8 A1.5 Aim and Purpose of TWGFIM ......................................................................................................................................... 8 A1.6 Criticism of Virtopsy ......................................................................................................................................................... 9 References ..................................................................................................................................................................................... 9 Lea A. Attias’s historical overview in Chapter C2 and in the introductory historical remarks at the beginning of Chapter D6 (both in this volume) demonstrate impressively that the Virtopsy Project rests on the broad shoulders of past researchers and their results. The historical aspects are also taken into consideration in this picture atlas, and this can be traced back to the illustration in Gil Brogdon’s basic text Forensic Radiology. [1] There, one finds the photographic image of what is probably the first examination of a corpse using x-rays in 1898. Then Brogdon spans an arch over the whole field of forensic radiology and toward the end of his book calls for a setting out toward new horizons: It is believed that forensic scientists in other disciplines would find radiologists in their area interested in cooperative efforts. Sharing of interdisciplinary skills and knowledge would improve the economy and effectiveness of investigative efforts, prevent some false starts and/or reinventions of well-worn wheels, and most important, expand scientific horizons.
When Brogdon wrote this the instrumentation that forensic medicine utilized for solving its tasks was already so advanced in its development that one really could not dare to think of embarking on new horizons in the forensic aspects of autopsies. In the 1990s, this was true for the development of photogrammetry, multislice technology in computer tomography, and the continuously improving resolution of the picture quality in magnetic resonance (MR) examinations. The technical mastery of the enormous flood of pictures and the possibility of postprocessing of the images also stimulated this optimistic mood. Then came the happy circumstance that an enthusiastic crew, interested in transdisciplinary
research, consisting of a core group from the University of Bern and supplemented by highly qualified international partners, could be found that was willing to “set sail for the open sea.” Michael Thali shows this development in “Biotope Virtopsy” (Chapter A4 in this volume). The crew knew that the goal on the horizon would not turn out to be a fata morgana, but rather that really new things were to be discovered. From year to year this conviction made them more and more curious. The project was given the name virtopsy, a combination of the words virtual and autopsy. The original ideas of digital autopsy, postmortem radiology, scalpel-free autopsy, or image-guided minimal invasive autopsy were discarded. In order to understand why virtopsy has been chosen one needs to consider the basic deliberations and views of the classical professional vocation of forensic medicine. Figure A1.1 can serve as a guideline to the professional program of an expert. The core business of every expert—it does not matter from which field—is based in the task given him or her by the judicature to report “findings and an opinion.” This is because an opinion must be founded on findings. These are usually then summarized as diagnoses or diagnosis groups. The findings thus form the objective foundation for a personal—that is, subjective—interpretation or an opinion. In this process, the report of the findings as well as the opinion must occur “according to best knowledge and conscience” of the Swiss Penal Code. The central function of the expert lies primarily in reporting the findings on which basis the specialist draws his or her argumentation. Through this the expert makes available the facts that are the empirical structure of the case that the court is to decide upon. In this function the expert complements the 3
© 2009 by Taylor & Francis Group, LLC
4
The Virtopsy Approach
Central Task of an Expert “Report of findings and opinion” Objective foundations
Subjective interpretation of the findings
Diagnosis = summary of the findings Findings
Opinion
Display of the findings
Determination of the findings
Documentation
Storage
“... According to best knowledge and conscience”
FIGURE A1.1 Diagram of a forensic expert.
expertise of the judge and is thus solely responsible for the correctness of the findings and opinion presented. The expert is obliged to transmit truth in reporting the findings. For this reason the expert must be highly objective. Since the findings are thus the actual source of a clear opinion, the expert must pay special attention to the display, determination, and documentation and storage of the findings. According to the contemporary gold standard, the display of the findings is attained through the technical procedure of sectioning—that is, the particular preparation artistry of the forensic medical expert and his or her assistants. The determination of the findings corresponds to the mental process involved in a differentiated, morphologic picture made by the naked eye or using a magnifying glass or a stereomicroscope. Not immaterially, it is dependent on the person’s own (i.e., subjective) optical memory. The documentation of the findings takes place through the written and photographic depositions of the findings. At the moment, how findings are stored is experiencing a transformation from the classical protocol archives to digital storage (Chapter B4 in this volume). Consider Gottfried Ephraim Lessing’s excellent definition of an objective judgment: “one can orient oneself, one that stands.” In light of this, the demand for objective reporting of the findings is understandable. This means, though, that an opinion should always be connected to the object or, in other words, oriented on the findings. Given this background, a uniform process is needed in forensic medicine in relation to the reporting of findings with these four elements: display, determination, documentation, and storage. This is, however, not always the case. Furthermore, as an example, the problem in the determination and documentation in contusion wounds will be treated. From these emerge diagnoses in a Babylonian language confusion such as Quetschwunde, Rissquetschwunde, and Quetschrisswunde (German for contusions and lacerations). The same is true, for example, for diagnostics in wound ballistics, where aufgesetztem Schuss, absolutem Nahschuss, or
© 2009 by Taylor & Francis Group, LLC
hochgradig genähertem Nahschuss (“shots at close range”) all refer to the same thing. This confusion arises from the fact that the diagnoses on which the opinions are based frequently lack a precise determination of the findings. Often the diagnoses already contain interpretations such as the concept of an absolute close-range shot, from which a conclusion, namely about the distance the bullet traveled, is derived. One must therefore always keep in mind the clear distinction between the findings and the opinion. This is because the reporting of the findings is a part of the objective knowledge of the specialist, while the opinion is a part of his or her subjective interpretation based on experience and evaluation of the objective findings. Nevertheless, the two elements of the expert’s activity are to be fulfilled “according to his or her best knowledge and conscience.” The real reason for the decision to realize the Virtopsy Project lies in the deep meaning of a clear distinction between findings and opinion because the enormous progress of the technological developments in the various fields of imaging has suddenly opened the possibility of reporting autopsy findings—both exterior and interior—more objectively than ever before. Thus, the idea of virtopsy as a minimally invasive form of autopsy was born. In order to explain this better, the contemporary, classical procedures for reporting findings are contrasted with those of the coming virtoptic approach. Fundamentally, it must not be forgotten that a forensic medicine reconstruction of a legally relevant event is only possible by means of an exact determination of the findings, and that from head to toe. This must be accompanied by precise geometric ordering and a spoken expression that is also understandable to laypeople. This must occur both at the location where the event took place and on the corpse. This procedure at the place of the event is called a local inspection whereas on the corpse it is known as an exterior and interior cadaver examination.
From Autopsy to Virtopsy: Oral Description versus Image: Value of Evidence
5
FIGURE A1.2 Preparation of a subcutaneous emphysema.
To describe the examination of the cadaver, three expressions with Latin and Greek origins can be employed: section, autopsy, and obduktion (the latter is German for “postmortem examination”). In the history of medicine, the term section appeared first. This seems understandable in that it describes the real, manual activity of opening the corpse: cutting, dissecting, and preparing. Later, mainly in French- and English-speaking countries, the term autopsy emerges and supposedly goes back to the historian Herodotus and means something like “observing for oneself.” Only at the beginning of the 18th century does obduktion surface (in German-speaking lands): The term means the “calm, prudent planning and ordered laying out of the findings”—hence the protocolling on an Obduktions-Zedel (“postmortem examination notice slip”), as it was called in 1706. Section, autopsie, and obduktion are in keeping with their meanings in relation to the concrete procedures involved in a postmortem examination, namely the preparation as a manual activity for the display of findings (section), an observation as the findings are determined (autopsy), and the written recording of the findings as a documentation of findings (obduktion).
A1.1 ABOUT PREPARATION This takes place using a scalpel and scissors. Only a portion of the body is examined, though, and thus many findings are left in the dark. This is because the sectioning technique has its limits when the results are displayed (e.g., for findings in the pelvic region). For reasons of piety (e.g., sections of the face in small children), one also maintains some distance. The display of the findings is thus always incomplete. They also depend on the person performing the examination
© 2009 by Taylor & Francis Group, LLC
because a proper preparation, and thus a clear display of the findings, always depends on the manual skill of the one performing the autopsy or the one making the preparation. A nice example is shown in Figure A1.2. Here something invisible is shown, namely air, accomplished by means of the most subtle, layered preparation. It was so cautiously prepared that the air bubbles, which one can recognize in the tissue, remained intact. Only if one is able to use the scalpel so precisely is it possible to display important crime-related details.
A1.2 ABOUT DETERMINATION OF THE FINDINGS This is the real mental process in reporting the findings. Do we in an autopsy—that is, by “seeing for oneself”–—really all see the same thing? The question is to be answered in the negative because the results of the observation—the findings on the corpse that are displayed—depend on a wide variety of factors. First are the strategic ground rules, a precondition for an exact observation, because as a play between “creating for oneself an overview and losing oneself in the details,” the process of observing is very complex. As an example, Figure A1.3 illustrates this play between overview and detail. At first one sees a separation of the skin’s continuity; looking more closely, additional, striped, radial, bluish alterations of the skin are visible. Schwarzacher [2] best describes this observation process: When we look at an object, this can take place in a variety of ways. Either we glance over it rapidly and attempt to obtain an overall impression or we make a systematic examination, point for point. The last approach, namely the systematic, step by step inspection, seems to be surer, but it has the danger that
6
The Virtopsy Approach
A
B
FIGURE A1.3 (A) A so-called patterned injury. (B) Injury-causing instrument (ashtray) matching patterned skin injury. the larger relationships are lost and one fails to see the forest for the trees. The most effective, therefore, is a right combination of both methods. One must mainly learn to see as naively as possible and nevertheless, like in an atlas turned to the right pages, to have all the old images available from memory. The better the separation and comparison a view towards the inside and towards the outside is successful, the more correct are the findings that are registered. In any case, one must be painstakingly exact and record all particularities. The love for detail distinguishes the master of his field. One also does not admit any valuations of the findings, i.e., diagnoses, in the protocol. This temptation lies very near, but giving into it will place the author in a bad light.
Observing this way, one will recognize the ashtray as the weapon, the bottom of which exhibits the striped, radial glass pattern that corresponds to the skin findings (Figure A1.3). In such concentrated, reflective, careful inspection of the findings also lies the reason for creating the forensic-optical memory for specialists. In the gold standard an objective determination of findings thus depends on high preparation skill and individually differentiated perception. With his characterization of perception as a mixture of brain, senses, and motor Heinz von Förster [3] describes these differing perceptions exactly. Gertrude Stein expresses the subjective part of perception with these words: “In the things of the mind, you will want the things you know.” Moreover, the dependence on adequate illumination, time pressure, and physiologic tiredness also has to be taken into consideration.
A1.3 ABOUT DOCUMENTING THE FINDINGS Documentation is the third step of reporting the findings. It is my opinion that it is the basis for a critical-scientific discussion for the interpretation of the findings. Only through the
© 2009 by Taylor & Francis Group, LLC
documentation of the subjective observations of the person performing the autopsy in terms of language, photography, or schemes does the first objective knowledge as the basis for an open, critical discussion arise. According to the criticalrational view (K. R. Popper), in this lies the foundation for a critical and free assessment of the opinion’s evidence by the organs of the judicature. Concerning the documentation of findings, Eduard von Hofmann [4], the founder of Austrian forensic medicine, made the following prerequisites: About the protocol to be made of the appearance, that is of the autopsy, it is to be written so specifically and comprehensively that it guarantees a complete and true view of the objects that were observed.
In order to realize this demand, it requires, first, a legal foundation and, second, a field-specific terminology. The legal foundations are extraordinarily variable. Austria, for instance, has an 80-page decree from 1855, while the Virchow “Regulativ” in Germany from 1875 comprises 25 pages, and Switzerland, with its 27 criminal trial regulations, has not even known such an ordinance. These were decided upon only in 1997 by the Swiss Society for Forensic Medicine within the framework of guidelines for postmortem examinations. The legal prerequisites for a comprehensive documentation of the findings thus differ from country to country. In addition, for an intersubjectively understandable documentation a field-specific terminology in the sense of a descriptive, formalized language that resembles a code is needed. This is because an autopsy protocol must be so written that later readers will be able to make an accurate picture of the findings even without having direct
From Autopsy to Virtopsy: Oral Description versus Image: Value of Evidence
perceptions. They should be able to see with their minds, so to speak. The persons making the postmortem examination should thus “photograph with words” in order to create the necessary conditions for a second expert or for their superior. For this a prerequisite is abiding by a language code that leads to understanding when it is read by a second expert. The autopsy made by the poet Friedrich von Schiller is put forward as a typical example of protocolled findings, whereby “in the left superior pulmonary lobe an irregular, indented, delimited wall covered with cheese-like, crumbly masses could be determined.” With this everyone who knows the language, or code, diagnoses an open pulmonary tuberculosis. We know now, however, that such a decipherable code, with which the findings are so registered that others (e.g., those providing a second opinion) are able to follow the ways of the expert, has been lost in our time. This will now be illustrated using the injury most frequently encountered in forensic medicine, namely the specific finding of a contusion. In Figure A1.4 the nine criteria of a contusion are listed. With this morphologic itemization it should be demonstrated that a diagnosis made from findings will be that much clearer the more one in the documenting keeps to the maxim “the difference makes the difference.” One no longer finds this differentiation of the morphologic general view in contemporary textbooks since the descriptiveformalized language for documenting findings has largely been forgotten. If one brings together the problems associated with the documentation of findings one sees that, besides the missing “from head to toe” description, the necessary language code is also lost. Not least, a three-dimensional displaying of the findings using language is extraordinarily difficult.
Findings of a Contused Wound ( ) splitting of the skin
irregular wound edge
hair root extending into the wound
wound angle not dried up
hematoma under the wound edges
formation of a wound pocket
drying of wound edges
tissue bridges in the depths of the wound
undermining of wound edges
FIGURE A1.4 Findings of a contused wound (Differentia spezifica).
© 2009 by Taylor & Francis Group, LLC
7
Reporting of Findings
Elements
Limitations
Presentation of findings
Not from “tip to toe”
Depending on the autopsy skill of the examiner Depending on postmortem changes
Compilation of findings
One only sees what one knows
Depending on the individual optical memory
Documentation of findings:
Loss of the formalized, descriptive terminology for describing the finding
Lack of time for the “tip to toe” description Different legal basis
Storage of findings:
Incomplete archiving
FIGURE A1.5
Limitations of the reporting of findings.
This is because for a description along the three spatial axes detailed, such comprehensive anatomic knowledge is necessary that, for the person performing the autopsy in the field of forensic medicine, is not readily available. Summarizing this information, the limitations of the reporting of findings according to the gold standard are shown in Figure A1.5. A. Werkgartner, full professor at the University of Graz in the last century, recognized early the documentation problem on the basis of a descriptive formalized language. He not only made the microscope acceptable in forensic medicine, but he also delivered the first paternity opinion based on blood group findings. It is thus not surprising that he was also the first to call for a correct photographic documentation. In his 1938 paper “Zur Bestimmung der stumpfen Hiebwerkzeuge aus dem Wundbefunde” [For identifying the blunt instrument from the wound findings], he stressed that the descriptions of skin wounds should always be supplemented by images: In order to make it possible to compare the forms of the instruments the wound is to be recorded in “normal projection.” It is sufficient for this purpose to apply three index marks on the edges of the injury area, roughly in the form of an equilateral triangle, to aim the camera at the center of the triangle and, by eye, align its optical axis so it is perpendicular to the wound area. It is advisable in such cases to maintain the distance between the front lens and the object that was calculated for the apparatus that is being used in order to obtain a known enlargement relationship.
With this idea, Werkgartner [6] intended to eliminate the perspective distorsio; in my opinion the first seeds of the Virtopsy Project—that is, autopsy in the three spatial axes— were planted with his publication.
8
The Virtopsy Approach
Philological background
Virtopsy–Improved Procedure
Autopsy (Herodotus): CT MR surface scan (scalpel)
– autos – self – opsomei – will look at
(self ) Observation
Virtopsy
Picture (protocol)
– Virtuel (franz.)/Virtualis (lat.) – better – opsomei – will see Telemedicine (objective discussion)
FIGURE A1.6 Philological roots of the word virtopsy.
A1.4 CONCERNING THE TERM VIRTOPSY The word creation virtopsy, as an amalgamation of the words virtual and autopsy, also has its valid philological roots (Figure A1.6). Over the course of history many words have altered their meaning, however, and this is true for virtual. If the scientific-philosophical mainstream employs the term virtual as an opposite to real and of this says that “virtuality breathes our subjectivity,” one can then understand the criticism concerning the term virtopsy from our own forensic circle of colleagues. It is rewarding for this reason to query the original meaning of the word in order to discover what this term really means. The roots of autopsy go back to ancient Greek. The term does not come from medicine, though, but from the humanities, namely from Herodotus, the historian. It is a combination of the words autos (“self”) and opsomei (“I will see”). With this Herodotus wanted to express that he wants to see for himself the historical places that he describes and writes about. Only later was the term used in medicine and, for determining findings, should express “looking at the findings.” Virtual appeared in the early 19th century, but not as an opposite to real. It was borrowed from French, which is derived from the Latin root virtualis and means something like “usable,” “industrious,” or “better.” Virtopsy does not, therefore, stand for a virtual forensic world that liquidates reality but for an improved, usable technique and better documentation for acquiring forensic medicine findings “from head to toe” objectively, meaning subject independent. If one now places the classic procedure of findings reportage against the virtopsy approach, one arrives at the following picture. Figure A1.7 shows the classic examination process as Classical Procedure Sectio technique
Dissection description, photographs
Autopsy observe for oneself
Expert opinion
FIGURE A1.7 The classic examination process.
© 2009 by Taylor & Francis Group, LLC
Expert opinion
FIGURE A1.8 Improvement of classic examination procedure by means of virtoposy
it is critically discussed. The cadaver would be opened using scalpels and scissors in order to display the findings. Through self-observation (i.e., autopsy), the morphologic details are determined, and finally in the postmortem examination protocol are documented as the basis of a critical-scientific discussion and interpretation (i.e., opinion). Figure A1.8 shows the improvement of this procedure by means of virtopsy. Scalpel and scissors are replaced by surface scanning, x-rays, and MR. Instead of a written protocol with decipherable, codified language, an objective picture arises—a picture with which the viewer is no longer left alone (i.e., looking at it alone) in his or her inspection and determination of morphologic details but that via teleradiological exchange can lead an objective discussion and can assemble an opinion, with more understandable forensic reconstructions for the court system. And so the worldwide assembled research leads to the field of virtopsy—maybe on the Technical Working Group Forensic Imaging Methods (TWGFIM) platform (http://www. twgfim.com)—and will open the door toward the goal of a minimally invasive autopsy.
A1.5 AIM AND PURPOSE OF TWGFIM A fundamental and definitive technological change in the field of forensic pathology (legal autopsies) and pathological anatomy (clinical autopsies) is emerging. Therefore, analogous to the fundamental technological change in forensic genetics, two decades ago the TWGFIM became necessary. This working group determines the scientific and technological conditions in order to achieve reliable and legally approved results with virtopsy and aims to promote an increasingly internationally standardized approach. TWGFIM offers a discussion forum for forensic pathologists and radiologists in order to exchange technical and scientific information regarding the virtopsy technology. The central aim of TWGFIM is therefore the validation of existing virtopsy methods and the regulation of respective instructions, titled “Guidelines for Virtopsy Methods.”
From Autopsy to Virtopsy: Oral Description versus Image: Value of Evidence
A1.6 CRITICISM OF VIRTOPSY The frequent critical objections to the virtopsy procedures and method, especially in relation to the transfer of the research results to the courtroom, are to be taken seriously. This is also understandable because the pictures that have served for illustration until now did not themselves constitute scientific knowledge. With virtopsy, though, computed tomography (CT) and MR pictures become the basis for a scientific opinion. This is not simply “old wine in new wineskins.” This is also why a critical analysis with the status of the scientific images is justified. One of the arguments against virtopsy is the false diagnoses in radiology. We know, though, that this mainly affects emergency radiology (i.e., the area that stands under time pressure). In forensic medicine the teleradiological exchange of the findings among specialists will surely be able to constrain this source of errors. The criticism of the image artifacts must be recognized as a scientific problem and must be solved. Virtopsy is also limited by image resolution. Currently, findings that could be essential for forensic interpretation (e.g., small, vital blood suffusions) have not yet been detected. Technological developments (e.g., 3 Tesla MRI) will also bring progress in this area. Naturally, the steps to multicolor image processing are also still necessary. These will make the replicability and understanding of the findings for laypersons (i.e., for the organs of the judicature) easier. The possibility that images can be manipulated has also been criticized. This is in the foreground because the judicature is increasingly confronted with the possibility of visual legal communication. Nevertheless, visual displays are strongly capable of being convincing in the courtroom. In contrast to analog photography where there is still a direct connection between the photographed object and the finished picture, the digital image no longer depends on light. Each pixel can
© 2009 by Taylor & Francis Group, LLC
9
be changed, moved, or deleted. If William Talbot, one of the inventors of photography, compared it with the true, unfalsified paintbrush stroke of nature, today the digital “brush” is used, which can generate and improve everything, but also falsify it. To solve this problem, corresponding claims for quality can justifiably be demanded. On the background of the coming transfer of virtopsy results into the courtroom—“virtopsy in the court”—still a final personal comment. Despite the impressive pictures that this atlas presents, language, indeed no longer—as in the protocol language—in its displaying function but still in its illustrative and explanatory functions, must be utilized. This is because only then will these impressive images be made precise and explained for the organs of the judicature. As M. Opitz [7] says, “Because it is language that contains all the other arts and sciences;” this is still true if in the change from autopsy to virtopsy “a new breed of high-tech detectives” arises [8].
REFERENCES 1. Brodgon BG. 1989. Forensic Radiology. Boca Raton: CRC Press. 2. Schwarzacher W. 1953. Befund und Gutachten. Wien Klin Wschr. 59:1−6. 3. von Foerster H, Poerksen, B. 1998. Wahrheit ist die Erfindung eines Lugners. Heidelberg, Karl-Auer-Systeme Verlag, pp. 15 24. 4. von Hoffmann E. 1898. Lehrburch der Gerichtlichen Medizin. Wien-Leipzig, Urban und Schwarzenberg Verlag, p. 28. 5. Hamper, H. 1964. Leichenoeffnung, Befund und Diagnose. Berlin: Gottingen, Heidlberg, New York, Springer Verlag. 6. Werkgartner A. 1938. Zur Bestimmung der Sstumpfen Kiebwerkzeuge aus dem Wundbefund. Beitr. Ger. Med. 14:66 97. 7. Opitz M. 1978. Buch von der Deutschen Poetery 1624. In: Gesammelte Werke. Kritische Ausgabe Vol. II Stuttgart, p. 347. 8. Watson A. 2000. A new breed of high-tech detectives. Science. 289:850 854.
A2
History of Virtopsy: How It All Began Michael J. Thali
Virtopsy originated in the mid 1990s. At that time, the Institute of Forensic Medicine of the University of Bern was quite involved and successfully active in DNA research. As things progressed, it became evident that the forensic DNA research milestones were in place. At that time, a request for a second revision process of a high-profile case in Switzerland was being considered at the Bernese Canton Appellate Court. Among other things, it dealt with the question of whether a ratchet wrench could have been the instrument that caused two specially shaped skin or bone injuries to a victim. In this case, which had occurred several years prior, the attribution of the ratchet wrench as the causal instrument had to be based on the skin injuries as seen in the photographic documents. A portion of the murder victim’s skullcap with an impression fracture that seemed thus shaped was still preserved. On the return flight from court proceedings outside the country, Dr. Walter Brüschweiler, director of biology of the Zurich City Police Scientific Services, and Professor Richard Dirnhofer, at that time the director of the Institute of Forensic Medicine in Bern, had the idea of trying to prove the attribution with so-called photogrammetric methods. Back in Switzerland they charged Marcel Braun, a technically experienced police official in the accident technical service of the Zurich City Police Department, with the task. Braun has been employed in the accident technical services for many years and is well acquainted with the photogrammetric methods utilized in documenting accidents (Figure A2.1). It was now necessary to transform the photogrammetric techniques, so to speak, from the macroarena of accident reconstruction to the microdomain of specially shaped injuries, or so-called patterned injuries. With a transfusion of funds from Bern to the Zurich City Police, Braun was equipped with the necessary technical instruments, and with them he began to transfer the technique of photogrammetry into the microdomain in order to solve the assigned task, which he was able to accomplish shortly afterward. In December 1995 I entered the Institute of Forensic Medicine immediately following my state exams, and in mid 1996 I had the opportunity to join the aforementioned group. The 3D visualization of patterned injuries in forensic medicine and the possible 3D documentation analysis for the inclusion or exclusion of possible weapons became the obsessive goals of our group over the next few years. A time of exciting, intensive, and pioneering work followed that often occupied us far past midnight and also filled our weekends. We were characterized by an open-minded spirit; the 3D
documentation technology developed concurrently and corresponded to our wishes, whereby over the years we were able to achieve the ambitious goals. In the beginning, the photogrammetric documentation was made using normal photo cameras. Various film sensitivities were utilized. Special projectors with special cross-hatching were developed by Braun (Figure A2.2 and Figure A2.3). Our autopsy hall had to be rebuilt. The joint venture between the Zurich City Police and the Institute of Forensic Medicine in Bern made it possible for courses to be conducted in which Swiss and foreign forensic specialists were trained. Among others, the first trials were made using pig heads from the slaughterhouse, and the injuries were documented using the techniques of that time and the reconstructions—they were all accomplished manually back then. It was possible to present papers at many international scientific conferences as well as to produce numerous publications. Soon the methods were also integrated into the scope of the forensic service. Somewhat later, the desire arose to gain some distance from the biologic trial materials that we had obtained from the slaughterhouse and to begin working with so-called synthetic anatomic models. At the Bernese Institute it was well known that very near Bern, in Thun, ballistics specialist Dr. Beat Kneubühl worked at the Swiss Ministry of Defense and had been occupied for several years with synthetic bones. It seemed quite reasonable, then, to contact him (Figure A2.4). At Dirnhofer’s request, Kneubühl made an appointment for us to meet with him at the Wittau-Matte shooting range. This visit formed the basis for a successful collaboration. In the years that followed, Kneubühl, an internationally recognized wound ballistics expert, intensified his collaboration with the Institute of Forensic Medicine, and since 1996 numerous further developments have been made in the area of synthetic anatomic models and their forensic applications, such as the development of the so-called scalp–skull–brain model and the development of synthetic vessels (Figure A2.5). The models consisted of individual components of synthetic skin and synthetic bone as well as simulated soft tissues. Gelatin and glycerine soap were used in the reconstruction of actual forensic cases. Attempts were made to reconstruct forensic cases with shots, blunt force, and stabs using these models. Happily, these were successful from the very start. These synthetic anatomic models were also used to document injuries with special shapes after they were made using photogrammetry (FigureA2.6). Since these synthetic models in terms of their behavior and characteristics approximated the features of biologic materials, it was possible from then 11
© 2009 by Taylor & Francis Group, LLC
12
FIGURE A2.1 Walter Brueschweiler (left) and Marcel Braun (right).
on to carry out reconstructions and photogrammetric documentations with these anatomic models. Injuries to anatomic models could now be documented in three dimensions using photogrammetry. It was thus possible from then on to dispense with biologic trial materials from the slaughterhouse. In 1998 we were able to perform a postmortem examination of a case with a gunshot wound to the head, which was not unusual for a forensic institute. The person only survived for a few hours at University Hospital Bern. Following the examination, the case was described briefly as usual at the morning report, and the accompanying computed tomography (CT) image folder from University Hospital Bern was also viewed. In the study of this collection of images, we noticed that all the findings seen at autopsy were also to be seen in these virtual radiology sectional images: the initial hemorrhaging and skin penetration in the region of the wound entry, the funnel-shaped broadening of the skull bone there, the blood-filled projectile channel with the scattered bone particles and air pockets into the brain, the funnelshaped broadening of the skull, and the skin opening where the gunshot exited the head. We were astonished as to how much could be seen in these sectional images. As a result we began to think about
© 2009 by Taylor & Francis Group, LLC
The Virtopsy Approach
FIGURE A2.2 The first new developments for 3D photogrammetric documentation in the 1990s.
whether to reactivate an idea that had existed since CT was invented— to attempt an image display of a corpse. The idea of postmortem imaging per se was not new. A study of the literature showed that a few individual papers already existed in this area. At a faculty meeting, Dirnhofer discussed the idea with Professor Peter Vock, director of the Institute for Diagnostic and Interventional Radiology at University Hospital Bern (Figure A2.7). In 1989 Vock, together with Willi Kalender, became the co-inventor of spiral CT, which opened the way for performing CT scans in 3D. Vock was immediately fond of the idea, and the decision was made to send me, the young assistant, to learn the radiology methods and applications in the radiology department at University Hospital Bern for two years. This was the beginning of forensic radiology at the Bernese Institute. In the beginning the radiological part was called “digital autopsy” or, somewhat provocatively, “scalpel-free autopsy” (Figure A2.8, Figure A2.9, A2.10, and Figure A2.11). Back then, external funds had to be acquired for the project because financial support from the institute resources alone was not possible. Eventually, using a digital autopsy flyer and an associated request we were successful in obtaining
History of Virtopsy: How It All Began
13
FIGURE A2.4 Beat Kneubuehl and a synthetic skull model.
FIGURE A2.3 The first new developments for 3D photogrammetric documentation in the 1990s.
FIGURE A2.5 Synthetic skull model—in the anatomically correct form and in a model form suitable for reconstructive use.
© 2009 by Taylor & Francis Group, LLC
14
The Virtopsy Approach
FIGURE A2.6 Application of blunt force to a skin skull brain model and following 3D analysis based on photogrammetry documentation.
FIGURE A2.7 Professor Peter Vock (left) and Professor Peter Dirnhofer (right).
© 2009 by Taylor & Francis Group, LLC
FIGURE A2.8 First publication of the virtopsy idea in 2000. In 28th AIPR Workshop: 3D Visualization for Data Exploration and Decision Making, W.R. Oliver, Ed., Proc. SPIE Vol 3905 (2000).
History of Virtopsy: How It All Began
15
FIGURE A2.9 Documentation of a patterned injury on a skull model using 3D photogrammetry and computed tomography.
FIGURE A2.10 Documentation of a patterned injury due to a blow with a hammer to a pig cadaver’s head using 3D photogrammetry and computed tomography.
© 2009 by Taylor & Francis Group, LLC
16
The Virtopsy Approach
FIGURE A2.11 Patterned injury of Figure A2.10: First 3D visualization and merging of the body surface and internal radiologic data set and subsequent matching with the injury-causing instrument.
several hundred thousand Swiss francs from the Gebert Ruf Foundation, which at that time specialized in the startup financing of projects that would not be supported through normal sources of research funding. After obtaining the external sponsoring, Bern University, in accordance with its practice, then contributed a matching amount to our research endowment fund. Thus the financing of the radiology portion of the project was assured. Under the direction of Dirnhofer and co-submitters Eva Scheurer, Wolf Schweitzer, and me, the foundation for the radiological virtopsy portion was in place. The “first generation” of the radiology part of what would later become the virtopsy team includes these names, with the addition of Kathrin Yen, who joined in taking up our task. A few years later, additional names appeared in the “second virtopsy generation”: Emin Aghayev, Christian Jackowski, and Silke Grabherr. The team was then completed in the “third virtopsy generation” by adding Stephan Bolliger, Steffen Ross, Lars Oesterhelweg, and Danny Spendlove. During the development of the radiology part of virtopsy, enormous technological leaps in the area of 3D surface documentation occurred. The classical photogrammetry was replaced by integrative systems. With the energetic support of the University of Bern, it was possible to later acquire two modern 3D optical surface scanners from the GOM Company. The technological implementation at the Bernese Institute was carried out initially by Braun and later was supplemented by engineers Ursula Buck and Silvio Näther. In addition, an intensified collaboration was possible with
© 2009 by Taylor & Francis Group, LLC
GOM for surface scanning and with Siemens for radiology scanning. With the turn of the millennium, further technological nuances and extensions were added to the project. Threedimensional surface documentation, imaging, and development of the synthetic anatomy model were supplemented with micro-CT and micro magnetic resonance (MR). In collaboration with the Institute of Medical Physics at the University of Erlangen and co-workers M. Karolczak, U. Taubenreuther, and Andreas Lutz, highly resolved displays of bone injuries using micro-CT became possible. During my fellowship at the Armed Forces Institute of Pathology in Washington, D.C., together with Kimberlee Potter and Bill Oliver, MR microscopy and micro-MR was introduced into forensic medicine. The group around Scheurer, together with Michael Ith and Chris Bösch, was able to implement the time-of-death estimation using MR spectroscopy. In addition, Aghayev, and later Ross, were able to perform postmortem biopsies. Grabheer, Jackowski, and Ross further refined postmortem angiography using intensive collaboration with the heart surgery team at University Hospital Bern, in particular through the participation of Erich Gygax and Barbara Sollberger. The whole project would have failed had we not had the support of the entire forensic institute in Bern. Besides the financial possibilities, the energetic support of Urs Königsdorfer, Roland Dorn, and Cosimo Carluccio was invaluable. There was always great support—often until late in the night—by radiology technicians Gabriel von Allmen, Elke Spielvogel, Karin Zwygart,
History of Virtopsy: How It All Began
17
FIGURE A2.12 Michael Thali presenting the idea of virtopsy.
Vreni Beutler, Suzanne Horlacher, Christoph Laeser, Carolina Dobrowolska, and Carole Stuker and by others in the Departments of Radiology, Neuroradiology, and MR-Spectroscopy. Enormous support, without which the project would have been condemned to failure from the start, was provided by the responsible district attorney offices of the counties of Bern, Aargau, Solothurn, and Wallis and the involved police corps and accident scene investigators, especially Hans Friederich, accident scene investigator of the cantonal police force of Bern. Following the millennium change, all developments were summarized under the term Virtopsy®, which stands as a label and the sign of scientific quality for all the utilized technologies. Later, following his retirement, Dirnhofer founded the Virtopsy Foundation, which essentially has the aim of collecting the financial means for the project, as well as the Technical Working Group Forensic Imaging Methods
© 2009 by Taylor & Francis Group, LLC
(TWGFIM), which forms the international guidelines for technological forensic applications. The Virtopsy Project, characterized by an open-minded pioneering spirit, has developed unusually well. The next years will be characterized by this further development (Figure A2.12). However, as with all new technological methods, caution should be exercised especially against misuse and low-quality applications. Efforts must be made to avoid allowing virtopsy and the applied methods to be discredited by unqualified application, as was once the case with DNA when it was a new technology in the 1990s and—due to an unsuccessful case—came to stand not for deoxyribonucleic acid but for “Do Not Analyze”! The co-workers of the Virtopsy Project will make every effort to always guarantee the highest demands for quality in forensic medicine and the utilized methods. To follow the past and future developments, please visit our website at www.virtopsy.com.
A3
Virtopsy ® after More Than Some 100 Cases: Statement and Perspectives of Forensic Imaging by Using 3D Optical and Combined CT/MRI Whole-Body Scanning Richard Dirnhofer and Michael Thali
CONTENTS A3.1 The History of Virtopsy, or How to Start Such a Project ............................................................................................. 20 A3.2 Advantages and Disadvantages; Resolution and Accuracy ........................................................................................... 21 A3.3 Evaluation and Validation of the Virtopsy Approach ................................................................................................... 22 A3.4 Replacement of the Inner and Outer Examinations (Classical Autopsy) ...................................................................... 23 A3.5 Costs and Grants ........................................................................................................................................................... 23 A3.6 Examination Times ....................................................................................................................................................... 23 A3.7 Practicality and Preview of the Implementation of the Different Techniques .............................................................. 23 A3.8 Quality and Added Values as a Virtopsy Aim .............................................................................................................. 24 A3.9 What Forensic Medicine Means to the Clinic and the Living Patient .......................................................................... 24 A3.10 Legal Issues and the Court System ............................................................................................................................... 24 A3.11 From Virtopsy to Virtobot? ........................................................................................................................................... 24 A3.12 Subdiscipline Forensic Radiology ................................................................................................................................. 24 References ................................................................................................................................................................................... 25
“The Forensic Autopsy Machine,” a paper presented at the 1992 American Academy of Forensic Sciences meeting in New Orleans (abstract number D49) by Dr. Gilbert E. Corrigan from St. Louis, contains this visionary statement: “[This] feasibility study is presented with a review of the currently available analytical techniques which may be applied to the human cadaver.” [1] At that time Corrigan’s vision was technically not to be realized as he wished, but he gave a preview of upcoming applications of imaging methods. Gil Brogdon, who summarized all the recent forensically relevant radiology techniques of the 20th century in his 1998 book Forensic Radiology, [2] provided more concrete visionary previews and guidelines: It is believed that forensic scientists in other disciplines would find radiologists in their area interested in cooperative efforts. The sharing of interdisciplinary skills and knowledge would improve the economy and effectiveness of investigative efforts, prevent some false starts and/or reinventions of well-worn wheels, and most important, expand scientific horizons. (p. 338)
From the forensic viewpoint the Virtopsy Group at Bern University began by evaluating and validating different
imaging modalities. Over the years, the Virtopsy Project (http://www.virtopsy.com) has developed into a multitool documentation and analysis research project consisting of the following: r 3D body surface imaging methods r Multidetector and multislice computed tomography (CT) r Magnetic resonance imaging (MRI) r Data merging of surface and radiological data r High-resolution micro-CT and micro-MRI (MR microscopy) r Magnetic resonance spectroscopy (time-of-death determinations) r Image-guided percutaneous biopsy r Postmortem angiography r Synthetic body model development The first results were presented at the beginning of the new millennium and later in a special “new imaging trends in forensic medicine” focus session at the 80th German Forensic Meeting in Interlaken in 2001, organized by the Forensic 19
© 2009 by Taylor & Francis Group, LLC
20
The Virtopsy Approach
Institute of the University of Bern. The basic publications followed in the years afterward. Virtopsy still focuses on these five forensically relevant topics: 1. Atrium mortis (i.e., a pathological reconstruction and explanation of the cause of death) 2. Relevant forensic patho-morphologic findings 3. Vital reactions 4. Reconstruction of injuries 5. Recapitulation and visualization It was possible for us in more than 80% of cases we have treated to diagnose the cause of death, based only on radiological data without autopsy. The Virtopsy Project met with worldwide interest: Scientists and journalists came to Bern, and invitations to come and speak were sent by, for example, many different universities, forensic institutes, the National Institutes of Health (NIH), the Radiological Society of North America (RSNA), and the U.K. Department of Health. Media echo was found on CNN and in the Wall Street Journal, the New York Times, the Associated Press, Popular Science, Spectrum der Wissenschaft, and other national and international journals and newspapers. Similar projects started all over the world through publications and personal communications, and terms such as scalpel-free autopsy, digital autopsy, virtual autopsy, computer-assisted autopsy, noninvasive or minimally invasive autopsy, necro-radiology, and image-guided radiology were created. As a response to our scientific presentations and publications, several questions were frequently asked. In addition to euphoric enthusiasm there was also some criticism, as is usual when a new technique is implemented in a field. After experience in hundreds of cases that underwent a combined 3D surface and CT/MRI whole-body documentation we would like to discuss the most important issues that have been sent to us by the scientific community over the past few years during article review processes and at scientific meetings. The goal of this chapter, then, is to answer these frequently asked questions. It is also intended to give other research groups guidelines and support for when they look to start a similar virtual autopsy project or are planning to buy some of the leading-edge technology used in the Virtopsy Project.
A3.1 THE HISTORY OF VIRTOPSY, OR HOW TO START SUCH A PROJECT Virtopsy®, the registered term of the Swiss Virtual Autopsy approach, is a research project initiated by Professor Richard Dirnhofer, former director of the Institute of Forensic Medicine at the University of Bern in Switzerland. It was carried out in close collaboration with the Institute of Diagnostic Radiology at the University of Bern, the directors of which
© 2009 by Taylor & Francis Group, LLC
are Professors Peter Vock and Gerhart Schroth. Under the direction of the University of Bern’s Institute of Forensic Medicine, over the years virtopsy has acquired a transdisciplinary and international dimension with worldwide collaborations in many medical disciplines (http://www.virtopsy. com). Before commencing such a project it is necessary to build a transdisciplinary collaboration among forensic medicine, pathology, and radiology. Divisions between specialties have to be bypassed, and the goal must be the sharing of interdisciplinary skills and knowledge. Only this scientific bridge and solidarity guarantee the morphological and radiological correctness and know-how. This transdisciplinary collaboration is one essential key factor of success. A new view of problem solving is only possible if conventional barriers are crossed over. Moreover, a functioning network based on technical and human resources is necessary. Good information transfer and transparency of all the involved institutes and persons are necessary to avoid and eliminate any criticism that is not relevant to the project and that only arises from poor communication. This open policy leads to an empowerment of all research project members. The integration of the forensic and radiological technician is necessary, and a key point of failure or success of such a research project relies on a good mix of experienced, older as well as innovative, younger people. The difference makes the difference: A good combination of human and technical or soft and hard factors is necessary. When a department does not have its own scanners, radiological scans are often only possible after hours in a radiological department. During postmortem scans it is more than useful if a radiologist and a pathologist with experience are in attendance so the first radiological diagnosis can be made before the autopsy. Our experience taught us that a doubleblind scientific approach is not always useful, because often a radiological finding, normally undetected during a classical forensic approach, has to be checked and vice versa. We can draw the conclusion, therefore, that our virtopsy cases were possibly some of the most precise, detailed, and bestdocumented autopsies to date. Philosopher Karl Popper’s [3] statement, “You only see what you know,” was demonstrated several times. Another relevant point is that a research project should never slow down the daily speed of the routine praxis; some delays occurred that had to be addressed and discussed with the justice institution or loved ones and relatives. In order to learn this new virtopsy approach and examination process and to gain experience, a relevant caseload is useful. As presented earlier, in forensics the following workflow is excellent: r 3D surface scanning to document body surface injuries r CT scanning to document any bone injuries and gross pathologies (takes only minutes and can be used as a screening tool)
Virtopsy® after More Than Some 100 Cases
r MRI scanning, with a tissue resolution that is higher than with CT in order to look at the tissue injuries r Conventional invasive autopsy r Comparison of the findings, following the wellknown Armed Forces Institute of Pathology (AFIP) idea in clinical pathology and radiology (see Radiographics Journal, a monthly publication) By doing this research work, an open-minded approach is necessary. It makes no sense to fear getting involved with it, and the virtopsy approach is not merely looking at images. Virtopsy aims at creating added value in forensic science. The goal of our approach is not playing around with some high-tech machines. The core point of the project is not to eliminate the classical approaches but to implement techniques in forensic medicine that are at the level of the current technology. As in the past, forensic science has to continue implementing newer techniques and to explore the extent to which new technology can be meaningfully utilized. During this process it is often not necessary to reinvent the wheel—it is often sufficient to modify some existing technologies (this was the same as with DNA). Quality at the highest level is the goal of virtopsy. Because the difference makes the difference, we registered virtopsy as a name, and it stands for a high-quality approach that is necessary in forensic medicine. Ideally, all the techniques would be located at the institute of forensic medicine. Due to the cost and radiologic knowledge, however, and the early phase of the virtual autopsy development, it is useful at the beginning to make the scans at a radiological department where the newest scanner technologies can be used and the radiologist and radiological technicians can provide their support and knowledge. It is important to use up-to-date technologies (e.g., multislice or multidetector CT scanners) due to their higher resolutions and faster examination times. Be aware that a project like virtopsy will generate thousands of radiological and autopsy photographic documents (it is often necessary to shoot dozens or hundreds of photos for further correlation with the radiology and autopsy findings). Until now, the best way to accomplish this has been demonstrated by the radiological department of the AFIP in their studies, which are published monthly in Radiographics Journal. The strategic and intellectual planning for taking pictures during autopsy that correspond to the topographic radiology findings is more than useful—it is the basic approach. So be sure and take the time to evaluate the data using modern software and workstations; avoid burying the data without drawing scientific conclusions and making scientific publications. It is necessary to include additional forensic techniques (e.g., histology, toxicological analysis, diatoms, microscopy) for validating the virtopsy approach. In addition, there is the option of making radiological organ examinations, for example, making the scans at higher resolution. Another possibility is to examine tissue samples at high resolution using micro-CT and micro-MRI. These will be the basic
© 2009 by Taylor & Francis Group, LLC
21
approaches or procedures that lead to a minimally invasive virtopsy approach in the future. Submitting papers dealing with forensic radiology was not easy at the beginning: The topic is not yet well known, and the theme is often too radiological in nature for a forensic journal and too forensic in nature for a radiology journal. It is the merit of some open-minded journals, especially the Journal of Forensic Sciences, that has given us the first platform for presenting our virtopsy feasibility studies. The reviews of the first Virtopsy Group papers were severe—often showing some lack of forensic or radiological knowledge by the reviewers of the complementary medical discipline. For example, we learned that the European term vital reaction is not very well known in the United States; radiologists, who deal clinically with living persons, were not aware of inner livores in the lung and often misinterpreted them as dorsal pneumonia. Virtopsy was thus a side-by-side learning process for the forensic and radiology scientists. Publishing began by using case reports and some specific basic or feasibility studies; subsequent publications will summarize case group studies and new technical approaches. Some of the new technical developments, such as postmortem biopsy and angiography, were based on postmortem animal studies. For example, in the MRI area we learned that the resolution and image quality depends on the body temperature during the scan, in contrast to clinical radiology where postmortem body temperature can vary greatly from normal body temperature.
A3.2 ADVANTAGES AND DISADVANTAGES; RESOLUTION AND ACCURACY There are still some limitations of the gold standard documentation, which consists of verbal descriptions, sketches, and 2D photographs. In contrast to the traditional field of forensic medicine, there have been enormous developments in the imaging and radiological fields. In the radiological examination, in contrast to clinical scanning, there are two relevant points: there are no respiration artifacts and no radiation limitation. As a result, the radiological pictures of the virtopsy examination are not limited or affected by these factors. In brief, the virtopsy approach has the following advantages: r Gives observer-independent and objective data archiving r Is nondestructive r Is minimally invasive r Provides actual-size documentation r Is in 3D r Is not necessary to touch the forensic evidence r Is not destructive of forensic evidence r Provides 3D geometry that is correct in xyz-axis or space documented
22
The Virtopsy Approach
r Real data-based; the basis for sound scientific reconstruction r Provides an alternative or additional examination tool for “difficult body area autopsy” (e.g., face, neck, pelvis) r Could be used in cultures and situations where autopsy is not tolerated by religion or rejected by family members (e.g., psychological reasons) r Provides ability to examine bodies contaminated by infection, toxic substances, radionuclides, or other biohazards (i.e., bioterrorism) r Provides 2D and 3D postprocessing for visualizing the findings by people not present during the examination r Gives greater understandability in court r Supports the process of quality improvement by digital archives (database for teaching, learning, education) r Teleradiopathology, teleforensic (“second opinion”) The virtopsy approach also has the following disadvantages (still present today): r Cost r Limited tissue resolution of radiology scanners r No visualization of organ colors (e.g., inflammation process) People frequently say that radiology did not show an autopsy finding or showed a finding that was not found during autopsy, and this argument is often used to demonstrate that radiology is not an accurate method. Be aware of misdiagnosis in radiology and results not obtained using the right technique or correct examination radiological protocol: This is often the reason for misdiagnoses. Misdiagnoses are possible in pathology, too. In retrospect, we can conclude that we have learned a lot by doing radiology and autopsy procedures in the same cases. It has sometimes occurred to us that by doing our virtopsy research, a diagnosis made by a radiologist led the forensic pathologist to a modification of the autopsy approach. It is certain that with today’s techniques, not all the findings are yet visible with clinical scanners, which is why we are performing additional postmortem biopsies and angiographies and working with high-resolution machines like micro-CT and micro-MRI, both of which provide microscopy-like information. An advantage of the project is the radiological documentation from head to toe. Normally such documentation is not done in a classical autopsy. Depending on the case circumstances, the experience, and the attentiveness of pathologists, they make the strategic procedure decisions and perform a toxicological drowning including diatoms assay, microbiological autopsy, or accident autopsy that includes the soft tissues of the back and extremities. 3D visualization is useful in bone trauma; the findings can be visualized in an unbloody manner, and the maceration process,
© 2009 by Taylor & Francis Group, LLC
which physically usually can take several hours or days, can be done virtually in some minutes at the computer workstation. Because there are no motion artifacts and radiation limitation in the postmortem examination, scans can be made at the highest resolutions (submillimeter). In addition, the radiological approach is useful in areas where an autopsy is limited by piety (face area) or in complexity (e.g., the pelvic area). The new multislice or multidetector CT generations only take some minutes for documenting a whole body. Today, if you are doing whole-body MRI imaging, which yields much better tissue resolution than CT, the main disadvantage is the long examination time. However, using new MRI software techniques (e.g., parallel imaging) the time required for an examination will decrease. Using high-field-resolution MRI scanners (e.g., 3 Tesla) tissue resolution will increase. Similarly, the visualization possibility will grow over the years using new computer programs and segmentation techniques, for example the coloring of special findings. It is possible, for instance, that the software will color a venous bleeding blue and that an arterial bleeding in red, epidural, subdural, and arachnoidal bleedings will be colored in a special way. An additional topic for investigation could be the automatic detection of findings, such as bleeding, infarctions, edema, and fractures. Today it is already possible to detect edema by MRI. To finally diagnose an inflammation process the virtopsy team is advancing methods like postmortem biopsy and angiography. The first results on animal and human studies showed satisfactory and encouraging progress. Virtopsy opens the way to a geometric documentation by merging both body surface and internal information together, leading to a geometric real data-based documentation, which is very useful, for example, in the area of vehicle accident reconstruction and research.
A3.3 EVALUATION AND VALIDATION OF THE VIRTOPSY APPROACH The Virtopsy Group is promoting the evaluation and validation of the virtopsy approach, but we are aware that virtopsy should not be advocated as a stand-alone procedure in the court system too early. 3D surface scanning is already accepted as a method in the Swiss court system, but more scientific research is indicated for the radiological portion. We are thus handling our cases like an accountant doing “double bookkeeping”: always a virtopsy documentation followed by a traditional autopsy procedure and documentation. It is a sign of great progress and also fortunate that several groups worldwide are now starting similar projects. Similar ideas are being promoted today by the NIH, which organized the “Non-invasive Autopsy” workshop in March 2003; by the AFIP in Washington, D.C.; by the Department of Health in London; and by other research groups in Sweden, Germany, Denmark, the United Kingdom, Australia, and Japan. It would be great if additional classical or clinical (i.e., nonforensic) pathology centers would do more postmortem research because there is still much to do (e.g., in the areas of
Virtopsy® after More Than Some 100 Cases
mors cardia subita; inflammation). Additional developments could even be postmortem molecular or genetic imaging based on postmortem angiography and other techniques.
A3.4 REPLACEMENT OF THE INNER AND OUTER EXAMINATIONS (CLASSICAL AUTOPSY)
23
funding from the Forensic Institute. A new step in the funding process is the creation of the Technical Working Group for Forensic Imaging Methods (TWGFIM; http://www.twgfim.com) and the Virtopsy Foundation (http://www.virtopsyfoundation.com). Collaboration between industry and the university will also be an important step in the future.
A3.6 EXAMINATION TIMES
In the beginning, the Virtopsy Project, with the implementations of the sectional imaging for documenting postmortem findings, was viewed by many as traitorous, even as a Judas kiss, to traditional, classical forensic medicine. From our point of view, however, virtopsy is a continuation of the centuries-old tradition of classical autopsy and only introduces new technologies into the field. It is a translation of findings (autopsy) into a new modality (radiology). This is not even all that new; the newest technologies are also being constantly implemented in forensic toxicology and also in the still young field of forensic molecular biology. It makes sense, therefore, to still perform classical, traditional autopsies in parallel with virtual imaging autopsies for a while and not yet implement virtual autopsy as the sole method too early in the court system.
Initially, examination times were quite long, especially when it came to 3D surface documentations and MRI scanning. With the further development of the technologies these time spans have been reduced enormously. A complete postmortem body surface documentation, depending on the resolution, now takes 30 to 90 minutes. One can have a full-body CT in 5 to 15 minutes. Developments in MR technology such as parallel imaging—the so-called total imaging matrix— have also reduced examination times considerably; today one can expect an examination time of 30 minutes. In addition, the further development of the so-called body surface coils and further software such as diffusion tensor imaging make a more detailed display of the findings possible.
A3.5 COSTS AND GRANTS
A3.7 PRACTICALITY AND PREVIEW OF THE IMPLEMENTATION OF THE DIFFERENT TECHNIQUES
Naturally, the newest technologies cost more than the classical examination instruments used in traditional forensic medicine. There is also little need to further discuss the fact that it is difficult to obtain research money and donations for forensic medicine because the social stigma associated with postmortem examinations does not especially encourage generous donations of research funds and donations. The enormous media attraction of forensic science recently, especially due to television series like “CSI: Crime Scene Investigation,” has not yet positively influenced generous sponsoring, unfortunately. In the research phase, depending on the case (e.g., cardiac infarction, gunshot to the head, complex pedestrian– vehicle accidents), the cost of a case handled by virtopsy is 2 to 3 times higher than normal postmortem examinations; however, it must be remembered that the cost in the starting phases of DNA research and implementation were 5 to 10 times higher in that field. Like with every new and not yet established research topic, acquiring money for starting such a project is very difficult. It was even hard to receive money from public organizations by mentioning the project efforts concerning public safety and medical quality control. Normally a lot of public money goes toward such work as heart and cancer research and stem cell research. Forensic medicine, a discipline already at the end of the medical hot topic list, often comes in last regarding obtaining financial support. It was very prestigious that our project was sponsored by a start-up grant by the Gebert Ruf Foundation and later through the medical faculty at Bern University as well as in-house
© 2009 by Taylor & Francis Group, LLC
Surface documentation consisting of photogrammetry and 3D opening scanning technique is today already state of the art at our university institutes. With this documentation method shaped wounds on bodies are documented both on the living as well as on corpses. With complex occurrences (e.g., criminal acts, accidents) the locations involved and the objects that are there (e.g., automobiles, airplanes, weapons) are also 3D and documented in close cooperation with the police. In the field of radiology it is quite likely that the classical x-ray methods in forensic medicine will be replaced with CT technology at forensic institutes in the coming years. Therefore, CT is an ideal screening tool for mass catastrophe management because by using a scanner like the ones used in the customs offices at airports, findings can be determined rapidly. In the field of MR technology, which is considerably more expensive, it will still take several years until this approach will be implemented in daily forensic praxis. The high-resolution micro-CT and MR technologies, which make a display down to a microscopic level possible, are still research tools today; until these are integrated into the daily praxis it will also take quite a while. The image-supported biopsy methods, on the other hand, will become established in the field relatively soon since they are relatively easy to manage. The main foundations of postmortem angiography are already in place, but further research is still necessary there as well. Nothing stands in the way, though, of implementing this postmortem angiography technique in the next few
24
years. With this the development of a “Virtobot,” in which all the aforementioned technologies are integrated into one instrument—in the sense of an all-in-one machine—is conceivable. Also important besides these technological—and in the end also software—developments, however, are also forensic virtopsy manuals that lead to an international standardization of the procedures. The TWGFIM represents a platform for this purpose.
A3.8 QUALITY AND ADDED VALUES AS A VIRTOPSY AIM As a whole, the virtopsy approach, grafted onto centuries-old knowledge and experience, leads to the increasing of quality and added value in the fields of forensic medicine and science. Similar to the earlier establishment of forensic DNA technology, important steps in the future will be that international standards are set up and that insufficient opinions and publications—by those jumping on the bandwagon—that are not technically of a high standard can be avoided. In Switzerland, the TWGFIM serves as a basis and container for such global standardization. The Virtopsy Foundation, grounded by Dirnhofer and similar to the Arbeitsgemeinschaft für Osteosynthese (“AO working group for osteosynthesis”) created some years ago by Dr. Maurice E. Mueller in the orthopedic scene, also has the task or the function of promoting the documentation of the research and teaching in the field of virtopsy. It is necessary that those interested in vir topsy in the future work together on an international basis at a high level. In the end, in light of the continuing growth of global terrorism, it should also be possible for the forensic field to acquire grant-based financing.
A3.9 WHAT FORENSIC MEDICINE MEANS TO THE CLINIC AND THE LIVING PATIENT According to our experience of postmortem imaging we can now transfer the technical expertise to living patients. This took place initially in the area of assessing the life-threatening, or mortal, danger that a strangulation victim went through. In a postmortem autopsy, which we have compared to postmortem imaging (in particular MR), we know how the findings in and around the neck after a strangulation look. As was true earlier, like in anatomy where anatomical knowledge is used for the clinic, today a transfer process from postmortem imaging to forensic clinical imaging is taking place. In the future, it will be possible thereby to document findings on the living that are invisible outwardly using modern imaging, and this will finally lead to a forensic opinion with more quality and added value. It is also known worldwide that in general the autopsy rates are declining in pathology. So the idea has arisen that this postmortem imaging could be used in place of the classical invasive autopsy. Since the autopsy rates have declined,
© 2009 by Taylor & Francis Group, LLC
The Virtopsy Approach
the NIH in Bethesda, Maryland, among others, also had the idea that one can possibly perform autopsies based on radiology and by this means could avoid having to make complete classical autopsies. Based on the virtual imaging autopsy pictures, targeted partial postmortem examinations could then be performed.
A3.10 LEGAL ISSUES AND THE COURT SYSTEM Before starting such a project, be sure to obtain all the necessary legal and medical ethical committee agreements. In a radiological department, the identities of those scanned have to be in the form of coded numbers. We scan our bodies in a specific double-bag system, created by our forensic technician, which provides identification protection and is also useful for hygienic reasons. The identification by a number and not by the name is useful for data security and provides a good basis for teleforensic medicine. It is very difficult to falsify the digital images encountered in forensic practice because a body scan consists of several hundred pictures. Thus, only highly trained professional imaging staff would be able to do it. Moreover, the real 3D-based documentations are imminently suitable for forensic animations. The documentation is objective; the data can be stored on a CD as evidence and are available at any time for a second opinion.
A3.11 FROM VIRTOPSY TO VIRTOBOT? In collaboration with industry partners we are developing a Virtobot, a system that will push the virtopsy idea to a Virtobot system and will combine tools like a 3D surface scanner, radiological scanners, postmortem angiography, biopsy, and high-resolution micro-CT and micro-MR.
A3.12 SUBDISCIPLINE FORENSIC RADIOLOGY The virtopsy approach has a great attraction for younger people who are planning to start an academic career in forensic medicine or radiology; every month we receive several requests from persons worldwide who want to start a fellowship or obtain a position at our institute in Bern. Besides the academic prestige, our project makes sense: Radiologists can learn a lot in the area of traumatology, since findings are confirmed or excluded by the autopsy. This may lead to the formation of a research organization for learning forensic radiology. That such a collaborative approach is useful is demonstrated by a picture published in a study in the well-known, high-impact journal [4]. This figure shows four intracranial bleedings. Looking through clinical-radiological glasses, the authors of that study overlooked the clinically uninteresting bleeding in the scalp tissue. By knowing about all five bleedings, it is possible for the forensic radiological doctor to conclude the axis of impact and the coup and contre-coup phenomena.
Virtopsy® after More Than Some 100 Cases
Additional newly detected topics of forensic importance are, for example, the following: r Gas embolism detection in, for example, the brain, heart-vessel system, organs of the abdomen, and soft tissue r Aortic collapse sign as indicator for fatal hemorrhage r Gunshot wound tracks r Differentiation of tissue damage r Tissue emphysema in strangulation cases At the beginning pathologists and technicians feared losing their jobs, but it now appears that the occupation is changing to a technology based on forensic tradition. Students are very interested in new technology, and the fame of forensic medicine is growing. The classic picture of bloody murder pictures is changing in the heads of the population, too. In the media virtopsy is frequently called scalpel-free or digital autopsy. Virtopsy is a multimodality method leading to an image-guided, minimally invasive 3D body surface and body interior documentation tool that now opens possibilities in forensic analysis and documentation. Because the multitool virtopsy approach will create a process of change in forensic medicine over the next decades, teaching will be an important and core topic over the next few years. Teleforensic medicine and the development of
© 2009 by Taylor & Francis Group, LLC
25
databases, therefore, are important; at the University of Bern we have already begun taking some steps in this direction. The model for the topics of documentation, education, and research is the orthopedic documentation system, demonstrated by the famous and highly respected Arbeitsgemeinschaft fuer osteosynthesefragen, working group on osteosynthesis topics (AO) and Maurice Muller Foundation /Switzerland, respectively. The TWIGIM, similar to the DNA Technical Working Group in the late 1980s, would be the promoter of such a development. Virtopsy is extending forensic medicine technically and measurably to a level like that of forensic toxicology and DNA. Virtopsy will lead to a new sort of “forensic detectives.” Let us conclude with a statement by the philosopher J.A. Schumpeter [5] “Innovation is the constructive destruction of tradition and old things.”
REFERENCES 1. Corrigan GE. 1992. The forensic autopsy machine. Abstract No. D49, AAFS meeting, New Orleans. 2. Brodgon BG. 1989. Forensic Radiology. Boca Raton: CRC Press. 3. Popper KR. 1974. Objective Erkenntnis. pp. 74 76. Hoffmann und Campe. 4. Matiello JA. Munz M. 2001. Images in clincial medicine. Four types of acute post-traumatic intracranial hemorrhage. N Engl J Med 344:580. 5. Schumpeter JA. 2005. Kapitalismus, Sozialismus und Demokratie. Stuttgart, UTB.
A4
Legal Side
CONTENTS A4.1 Virtopsy and the Law ...................................................................................................................................................... 27 A4.1.1 Synopsis ........................................................................................................................................................... 27 A4.1.2 The Objectives of Postmortem Investigations ................................................................................................. 27 A4.1.3 Compulsive Autopsies ..................................................................................................................................... 28 A4.1.4 Question of Consent ...................................................................................................................................... 29 A4.1.5 The Standard of Proof ..................................................................................................................................... 29 A4.1.6 Religious, Cultural, and Other Objections to Invasive Techniques ............................................................... 31 A4.1.7 Some Other Consequences of the Use of Noninvasive Techniques ................................................................ 33 A4.1.8 The Court Process ......................................................................................................................................... 34 A4.1.9 Conclusion ..................................................................................................................................................... 34 A4.1.10 Notes ................................................................................................................................................................ 34 A4.2 An Extraordinary Death .................................................................................................................................................. 34 A4.2.1 References ........................................................................................................................................................ 36 A4.3 Virtopsy and the Swiss Legal System: New-Evidence Law in Forensic Medicine? ....................................................... 36
A4.1 VIRTOPSY AND THE LAW Graham P. Segal, OAM[1]
A4.1.1 SYNOPSIS All societies have a legitimate interest in the reason for the death of their citizens. Different societies will perceive different interests in the death process, and indeed those interests will vary from time to time within the same jurisdiction. The interests will vary depending upon social, political, and scientific changes that take place within each jurisdiction. This section examines these issues and the legal consequences that arise. It is contended that the introduction of noninvasive techniques will have a significant impact on both the social issues and the law.
A4.1.2 THE OBJECTIVES OF POSTMORTEM INVESTIGATIONS Most jurisdictions have developed a set of legal principles that will govern the circumstances in which and the means by which it will inquire into the death of its citizens. Inevitably, the legal framework will need to accommodate, and commonly does accommodate in different ways, the consideration that will govern the investigation by autopsy or otherwise depending upon the purpose for which the process is to be undertaken. It is therefore essential that, in considering the legal environment in which the autopsy process takes place, one must first recognize the various social imperatives that dictate the need for such an examination. The recognition of the particular purpose is essential to framing appropriate legal
principles that will be cognizant of all of the interests that are involved in the process. Moreover, inevitably, interests will conflict, and a legal system will need to have a process that recognizes that potential conflict and provides for a means of the resolution of that conflict where it arises. The purpose of an autopsy examination is essentially to determine or otherwise assist in determining the manner and cause of a person’s death. That statement, it is suggested, while correct is inadequate to define when such an examination should take place and the manner of the examination. At the outset it should be noted that the autopsy is but one of the steps that can and is taken for the purpose of determining the answers to both questions. Moreover, it is to be recognized that the autopsy process is not always necessary to answer either question. It is, however, commonly a very important means of answering those questions and, on some occasions, the critical means by which the questions are answered. In their text Death Investigation and the Coroner’s Inquest, Ian Freckelton and David Ranson [2] identify the aims of autopsies (forensic and clinical) to include the following: r Confirmation and determination of the identity of the deceased r Identification of the injuries and natural disease r Reconciliation of events in life with the presence of anatomical and pathological features r Determination of the extent of injuries and natural diseases r Evaluation of the effect of medical treatments r Assessment of the mode of death 27
© 2009 by Taylor & Francis Group, LLC
28
The Virtopsy Approach
r Determination of the cause of death r Comprehension of the mechanisms involved in the death r Provision of an educational resource for the medical profession r Provision of tissues for use in medical research and therapeutic procedures r Retrieval of trace evidence and other samples for use as evidence in court r Reconstruction of the circumstances surrounding the death It is suggested that the societal interests in the autopsy procedure might be identified as follows: The determination of whether the death may have been occasioned by a criminal act The obtaining of evidence necessary to either successfully prosecute the person who may have committed a criminal act or to exonerate a person who might otherwise be suspected of causing the death by a criminal act The identification of the deceased The provision of an educational resource The obtaining of evidence necessary to enable the determination of potential civil litigation including compensation to the deceased’s family The determination of the existence or suspicion of negligent medical treatment The maintenance of quality control within the hospital and other areas of health-care practice Assisting in the prevention of industrial and other circumstances of accidental death with a view of reducing the incidence of death and injury The determination of genetic conditions for the benefit of the family of the deceased and subsequent generations The determination of means of reducing death and injury occasioned by wartime activities including the effectiveness or otherwise of measures used to prevent death and injury The general increase in medical and scientific knowledge The provision of comfort to the family of the deceased in understanding the reason for the loss of their loved one It is apparent that each of those objectives carries with it a social benefit. It is accordingly appropriate that a legal system recognize those benefits. However, this is not to say that the mere possibility of one or more of those benefits being obtained by an autopsy process should, without more, be sufficient to mandate that that process be carried out particularly over family objection. The particular benefit with respect to any of the objectives noted may in a particular case be minimal. Moreover, in a particular case, there may be other
© 2009 by Taylor & Francis Group, LLC
countervailing interests at work such as cost and manpower. Further, the cause of death may be apparent without any invasive procedure, but the manner of death may be a matter requiring significant investigation and yet may not require any form of autopsy.
A4.1.3 COMPULSIVE AUTOPSIES Commonly, forensic autopsies occur within a compulsive legal environment. Of course, compulsion need always operate. Indeed, a family may request such an examination. In Australia, Young CJ in Eq [3] spelled out the various parties who had the right, according to the common law, to the possession of the remains of the deceased. He pointed out that common law recognized the superior right of the coroner in this regard in circumstances where the law otherwise gave the coroner jurisdiction to investigate the cause of death. The law that permits or mandates an autopsy examination must therefore be seen to operate as an inroad into the rights vested in such other persons having such rights as might be conferred by the appropriate law of any particular jurisdiction. Superimposed upon the right of the deceased’s representative is the right of the coroner or other equivalent officer to require a particular form of postmortem examination [4]. In New South Wales (NSW), the Anatomy Act [5] dictates the right of the deceased to determine that his or her body will be used for scientific purposes. The Human Tissue Act [6] and the Coroners Act [7] dictate the circumstances in which human organs and tissue can be removed, used, and retained. The three different enactments are reflective of the fact that postmortem, human remains may have different purposes. There is always the risk of those purposes being confused, often with significant consequences when the examination is invasive. It was thought at one time at the NSW Institute of Forensic Medicine that, once the coroner had ordered a postmortem examination to be conducted, it was appropriate to take steps for scientific or educational purposes. A government inquiry soon demonstrated that to be wrong [8]. Quite apart from circumstances of legal compulsion, the persons with the right to determine—in accordance with the principles to which I have referred—the manner in which the body of the deceased will be disposed may well agree to an autopsy examination. Questions arise as to the disposition of the body parts taken in the course of that examination. Such body parts may be taken because their examination may be legitimately necessary for the purpose of determining the manner or cause of death. Moreover, it may be necessary that such body parts be retained for the purpose of being used as evidence in subsequent proceedings. A legal framework is needed to determine when such body parts need be retained, the manner of their retention, and their disposal following the completion of the purpose for their retention. As will be seen, one of the advantages of radiological techniques is the fact that retained images may give rise to a lesser need to retain organs for this purpose.
Legal Side
A4.1.4 QUESTION OF CONSENT Different jurisdictions provide criteria that determine when an autopsy procedure may be carried out both with and without consent. Further, provisions are made as to who may give that consent. Generally, in a legal system regimes will be found that are compulsive in some circumstances but in which a right of objection is grafted onto that system. In some jurisdictions, the consent might only be given by the deceased during his or her lifetime. The right to object may be given by the legal personal representative, yet in others, it may be given, for example, by a statutorily defined next-of-kin [9] or a more broadly based category of person [10]. Disputes may arise between various persons seeking to assert their position in respect to the remains of the deceased. This type of dispute is essentially beyond the scope of the present discussion. However, the problem may be real when those authorities having to determine the exercise of a compulsive power must consider the views of various family members and may well be confronted with competing and even antagonistic positions between, for example, family members and other persons claiming a relationship with the deceased. In New Zealand, the coroner may identify a small number of a larger group to represent the objecting interest. The compulsive power exercised by coroners, medical examiners, and other authorities such as police in different jurisdictions must be viewed at two levels. The first is to understand and define the circumstance in which that power may be exercised. The second is to consider the manner in which that power might be exercised. Citing the New South Wales example and indeed the circumstance that in general terms exists in all states within Australia, the position is broadly that the statute in force in each state and territory of that country specifies the circumstance in which the coroner (being the appropriate official) has the power to order both a coronial investigation and inquiry (an inquest) and, in consequence of the vesting of that power, the power to order a postmortem examination. The systems in operation through the various countries of the world vary considerably. In some countries, the position is different from state to state (e.g., United States or from Canton to Canton, as in the case of Switzerland). Similarly, the criteria in which deaths will be investigated by the appropriate functionary vary from country to country and from different jurisdictions within various states. It is probably fair to say that the functionaries involved in the decision-making process vary among judicial, administrative, medical, and police functionaries. Criteria giving rise to the compulsive power, while varying significantly from one jurisdiction to the other, commonly involve notions such as natural cause of death, suspicion of criminality, and, less often, the possible involvement of medical negligence. Appended is, by way of exemplification, a list of the circumstances in which the power exists in New South Wales, the United Kingdom and Israel. The general thrust of the
© 2009 by Taylor & Francis Group, LLC
29
circumstances in which the compulsive power exists is similar in other common-law jurisdictions. Issues may arise as to whether, in a particular case, the circumstances described may exist. Beyond that, the courts in Australia have demonstrated that proper consideration must be given to how those powers are exercised. The manner in which the power is exercised has been tested in the judicial framework where objections have been lodged to the process by family members [11]. In Deitz v. Abernethy, the Court of Appeal of New South Wales denied that the power asserted by the coroner was in effect unfettered. The court held that the coroner was duty bound to take into account, in exercising his discretion to order an autopsy, the religious and other sensibilities of the family of the deceased. The legislative as well as judicial development of these concepts is discussed later in the chapter.
A4.1.5 THE STANDARD OF PROOF When assessing the value of any particular technique in forensic investigation, consideration must be given to the appropriate standard of proof in the context of the object of the examination. When regard is given to the various circumstances in which a postmortem examination might be undertaken and the various purposes for which it might be undertaken, it becomes apparent that proof of particular matters will be seen from a different perspective and will involve different considerations depending upon the particular purpose. In considering a finding a coroner ought to make as to the manner and cause of death, the degree of satisfaction required in the Australian context is the balance of probabilities (i.e., which fact at hand is more probable). A coroner may make recommendations to the government arising out of the exercise of coronial jurisdiction. In considering the circumstance of a road accident case and public issues that might arise, a coroner may feel compelled to make a recommendation if he or she is satisfied that the matter, which might give rise to the recommendation, is sufficiently demonstrated by the particular case being considered and the extent to which the circumstance might be at risk of repetition. In deciding whether to make such recommendations, legalistic notions of standard and onus of proof necessary to determine the rights of parties may not be appropriate. To take an example, some years ago I was involved in a case of a young girl who fell off a bus and was dragged behind it without the driver being aware of what was happening for some considerable distance. The child was small, and it appeared that her foot had become caught in the doors of the bus as it closed while she and her mother were trying to alight from the bus. The accident demonstrated that the bus driver was confronted with a blind spot in his rear-view mirror so that the child stood too low to be able to be seen through those mirrors.
30
In consequence of the coronial inquiry, evidence was called to demonstrate not only the reason why the event occurred but also the matters that needed to be addressed in terms of the mirror systems of the bus and the manner in which passengers standing in the bus might obliterate the view of the driver so that systems could be devised to avoid a repetition of that event. The invasive postmortem that occurred played no part in determining the issue in the case. For the purpose of the subsequent prosecution of the bus driver, the causal relationship between the mirror system and the injury to the victim was critical and was required to be proved beyond reasonable doubt. For the purpose of the coroner’s recommendation, different considerations apply. If an investigation is being carried out in the context of a circumstance that may lead to a criminal prosecution or where one might wish to exclude the possibility of a particular individual being responsible for a death, very different considerations apply. In this context, at least in the commonlaw countries, the proof of guilt must be beyond reasonable doubt—that is, the onus is upon the prosecution. Accordingly, if the defense can point to any circumstance that might reasonably be consistent with the hypothesis of innocence, the accused person is entitled to an acquittal [12]. Moreover, if the prosecution failed to, as it were, discount another cause of death, the accused is, in many circumstances, entitled to an acquittal. That type of circumstance (although the coroner does not sit to determine guilt or innocence) then must dictate a means of investigation that does not leave open the prospect of a cause of death other than what is asserted against an accused person. It is in this context that a partial autopsy may, in some cases, be fatal to a subsequent criminal prosecution. Moreover, it will be apparent that autopsy examinations undertaken for any scientific or teaching purpose involve no such consideration but are dependent upon the relevance of the particular examination to the scientific objective being undertaken or the particular educational purpose sought to be achieved. In circumstances relating to the identification of human remains, it needs to be understood that a great deal, both emotionally and legally, will depend upon that identification. Families are entitled to the utmost care in the conduct of any examination so that an incorrect conclusion with all that will entail will not result. In considering this question of standard of proof, recall that the autopsy does not in itself represent the only piece of information upon which reliance can be placed for the purpose of answering the relevant questions. Accordingly, in considering the importance of any (if any) additional information that might be gained from an invasive process, there needs to be brought into account the evidentiary value of information available outside of the autopsy process—such information commonly being clinical history. Within the common-law world, by and large two, or possibly three, distinct standards of proof, namely, balance of probabilities—applicable in civil cases and beyond reasonable doubt—are applicable in criminal cases. Such concepts have little to do with science or medicine. Accordingly, when an investigation is undertaken or any step in the autopsy process
© 2009 by Taylor & Francis Group, LLC
The Virtopsy Approach
is taken for medical or scientific purposes, the question of accuracy or standard of proof required depends entirely upon the purpose of the scientific investigation. In the context of the rights of parties inter se in civil cases, the consequences for the parties are monetary and not penal. There is yet a third category, namely, that of professional consequences that may arise by reason of the intervention of professional disciplinary bodies. At least in Australia, the courts [13] have determined a hybrid standard of proof: comfortable satisfaction. In such cases, a health-care professional’s reputation or even livelihood is at stake. In the case of the criminal law, a person’s liberty is often at stake, let alone his or her reputation. It should be noted, however, that this approach is really a proper application of the civil standard in cases where one would not likely expect, because of the seriousness of the matter, the conduct of which complaint is made [14]. It is a guide as to a judicial method in resolving factual issues [15]. The forensic pathologist must be conscious of the particular circumstance in which the process is carried out and the potential consequence of that postmortem examination. Often the forensic pathologist will not know, particularly at the early stage of the legal process at which he or she is involved, where the future of the case may be headed, but in other cases it will be tolerably plain. Regarding the latter, the level of satisfaction may be adjusted accordingly. In a case that may have potential criminal consequences, it would be unacceptable to leave open the possibility of an innocent person convicted or a guilty person acquitted by failing to exclude possibilities as to the cause of death that the most detailed investigation might reveal. For that reason, virtopsy provides the prospect of enabling a larger-scale screening process that may help determine the extent and direction of future medical investigation. However, it is a mistake to think that every death needs to be explained in a manner or to the extent such matters of public importance are at issue. The former chief judge of the Common Law Division of the Supreme Court of New South Wales, Justice Wood, in Krantz v. Hand [16] (more fully discussed in the following section) had to consider the extent to which knowledge of the cause of death may be of importance having regard to the availability of other factors that sufficiently demonstrated a death by natural causes. The decision in Krantz v. Hand is significant not only for its affirmation of the importance of cultural and personal considerations of the family of the deceased—a topic to which I will return—but also for its recognition of the proposition that the standard of proof required of a coronial investigation may depend on the surrounding circumstances. It may well not be necessary to determine which of two or more processes lead to a death. This case also causes one to focus on the important distinction between the concepts of cause of death and manner of death. The more the evidence tends to establish that the manner of death is not due to criminal activity or other matters of public significance, the less important it is to demonstrate a cause of death—although on some occasions it may be necessary to investigate the cause of death to determine manner of death.
Legal Side
When matters concern issues relating to public health such as quality control within a hospital system, aside from any civil rights that may arise as a result of medical misadventure in a particular case, the matter has to be determined in accordance with the scientific method. Even that scientific method may be something that gives rise to different standards of satisfaction. For example, a circumstance may arise such as a step in infection control which indicates that a serious consequence in terms of danger to the public may occur if a medical precaution having little or no negative effect is not taken. In such a circumstance, a relatively low standard of proof of medical event or its consequence may be adequate. It follows that, in considering the need for and the extent of any autopsy process, the purpose of that process and the reliance that may be placed upon its findings must be kept steadily in mind. Moreover, there is a need to also have in mind whether a particular autopsy will, irrespective of its outcome, have any significant effect in terms of the objective sought to be achieved. The process leading to death is, of course, commonly complex. Some aspects of that complexity may have relevant forensic implications, and others may not. In undertaking any form of postmortem examination for a forensic purpose, the importance to the objective of the examination of the particular process has to be considered. That factor especially has to be considered when the process and all that is entailed in it have to be weighed against other social needs and objectives including the distress of the family as well as the allocation of increasingly scarce medical resources. For many years, it has been said that the standard threecavity autopsy is the gold standard of investigation. Whether that is a meaningful statement is, in the light of the aforementioned, open to question. Any attempt at comparison between the three-cavity autopsy and virtopsy by reference to the former being the gold standard is simplistic and certainly uninformative.
A4.1.6 RELIGIOUS, CULTURAL, AND OTHER OBJECTIONS TO INVASIVE TECHNIQUES Elsewhere in this work, the reader will find discussions demonstrating the extent and basis of objection by particular groups within society to invasive techniques. Those objections are to be found among many groups. With the globalization of society and the increased movement of populations, many societies are now composed of a variety of ethnic and religious groups. The consequence is that many societies, if they are to accord to all of their citizens freedom to express their religious and cultural practices, must find a way of balancing those expressions and other important social objectives. The autopsy process has proved to be a significant instance of such cultural and social conflict. That conflict has sought to have been ameliorated, if not resolved, through legislative, judicial, and bureaucratic measures. Some attempt has been made to suggest that the judicial process that has been brought to bear on this issue has been
© 2009 by Taylor & Francis Group, LLC
31
no more than an attempt by the courts to consider the psychological impact of cultural beliefs rather than the substance of the cultural objection [17]. However, such an approach not only denies the words that came from the various judges but also is likely to lead to the error of requiring proof of psychological damage before any objection made to a court can be upheld. As will be shown, it is quite plain that law reform bodies, the courts, and bureaucratic decisions focus on the respect that society should accord to people’s sensitivities and their need to observe their religious and cultural systems of belief and practice. Deitz v. Abernethy, mentioned earlier, was a case decided before any statutory right of objecting to autopsy was established by legislation. In that case, Mr. Deitz, whose eyesight was failing, was struck and killed by a bus. He was hospitalized but died shortly thereafter. The manner of his death was evident from the surrounding circumstances. The injuries he sustained were substantially identified in the hospital prior to his demise. That hospital examination included a computed tomography (CT) scan, which demonstrated a subarachnoid hemorrhage entirely consistent with the impact of the bus upon Mr. Deitz. At the hearing of the application, counsel for the coroner sought to object to the treating neurologist, giving evidence upon the basis that it did not matter what the neurologist had to say because the coroner’s decision was, in effect, a matter for him. The court rejected that contention and admitted that evidence. Ultimately, the court concluded that the autopsy ought not to proceed and that the decision of the coroner was open to review; upon review, the coroner’s decision was found to be, as a matter of administrative law, unlawful because of its failure to be properly exercised. The decision was improperly exercised because the coroner had failed to regard the medical evidence that established the probability of the cause of death and the religious beliefs of the deceased’s family—a belief system shared by the deceased in his lifetime. In coming to his conclusion, the trial judge, Abadee J, adopted the words of Beach J in Green v. Johnston [18] as follows: In a multicultural society such as we have in this country, it is my opinion that great weight should be given to the cultural and spiritual law and practices of various cultural groups forming our society and that great care should be taken to ensure that their laws and practices, assuming they are otherwise lawful, are not disregarded or abused.
The trial judge continued that theme by adding the following: What Beach J said can be thought to be applied equally to people’s religious beliefs as well as their cultural and spiritual laws. I do not believe that public interest in a case like this demands that further grief or anguish or hurt be caused to the plaintiff and to members of her family.
The decision was upheld in the Court of Appeal. Shortly after the decision in Deitz v. Abernethy, the Parliament of New South Wales enacted legislation
32
The Virtopsy Approach
similar to what had then been recently enacted in Victoria (the state in which Green’s case had been decided) providing for a right on the part of families to lodge an objection to an autopsy with the result that, if the objection was not upheld, the Supreme Court could conclude that it was not appropriate in the circumstances to order such an autopsy. The effect of that legislation, which is now to be found generally in the laws of the Australian states, is that the Supreme Court’s power is enlarged so that a Supreme Court judge can effectively make his or her own decision in lieu of that of the coroner without the need to show that the original decision was unlawful in the manner already described. In addition, the legislation in each case provides a procedure to be followed for the making of an objection and provides in each case a system whereby the coroner’s decision to conduct the autopsy is restrained until the determination of the application by a Supreme Court justice. The first case to be decided in New South Wales after the introduction of that legislation was Krantz v. Hand [19]. In that case, Mrs. Krantz, 84, was found deceased in her bathtub when her son came to check on her. Mrs. Krantz did not like seeing a doctor and was therefore a person who fell within one of the categories of cases attracting coronial jurisdiction, namely, she had not seen a doctor for three months prior to her death and, accordingly, no doctor was allowed under New South Wales law to write a death certificate. The police reported no suspicious circumstances. Her premises showed no signs of forced entry. Mrs. Krantz had bruising and abrasions consistent with a history of falls, which was otherwise demonstrated by the evidence. She had exhibited in recent days a decline in her health including shortness of breath. A cardiologist gave evidence that the symptoms (not fully described in this chapter) were such as to make it likely that Mrs. Krantz suffered a heart attack, although no certainty could be expressed in this regard. The court concluded that it was correct to say that the cause of death could not be stated with any certainty. Nonetheless, the court held that whatever the cause of death may have been, there was no sufficient public interest in identifying that cause with any particularity as it gave rise to no criminality and no other particular social benefit of significance could be pointed to by the coroner. The suggestion that advancement of medical knowledge might be improved by the collection of information was found to be an inadequate basis to override the family objection in the circumstance. The fact that the court did not conclude that the cause of death had been established is important. In the course of his judgment, the trial judge, Wood CJ at CL, adopted the words of Beach in Green’s case, Wood said the following: I express my entire agreement as to the appropriateness of taking into account religious beliefs of the family of the deceased where they can be demonstrated to be both genuine and to accord with the faith of those concerned.
© 2009 by Taylor & Francis Group, LLC
Wood went on to say [41]: That is a matter to be taken into account although it will not necessarily be determinative in any given case. In some circumstances, it may be that there is evidence pointing to foul play, which would need to be investigated in order to ensure execution of the due process of the law. In other circumstances, there may be evidence of a possibility of an outbreak of serious infection which would need to be investigated in order to cater for the public interest. Additionally, there may be cases where it could be in the interests of the immediate family of the deceased to determine whether there is some genetic predisposition to serious disease that might possibly be treated or detected in its early stages if the possibility of its onset is known.
Further, Wood said the following: I can see no possible public benefit in determining which, if any of those events brought about death or indeed whether she suffered from some occult malignancy.
Re: The Death of Unchango (Jr); Ex-parte Unchango (Sr) [20] was a case of a sudden infant death of an aboriginal child. The postmortem was objected to on religious grounds. The objection was upheld by the West Australian Supreme Court. The father claimed the desecration of the remains of his child was contrary to aboriginal custom and faith. Again, the court concluded that the death was brought about by natural causes: Whilst there is undoubtedly potential for a post-mortem to reveal the death may not have been caused by Sudden Infant Death Syndrome but by a natural cause such as infection, nevertheless that would not really advance the matter a great deal in circumstances such as this. One should take into account the very strong cultural beliefs held by the relatives and by the community at Kalumburu and the effect the post-mortem would have on them by way of emotional trauma particularly in view of the fact that it would prohibit, in their view, the spirit of the deceased remaining in the body and returning to the body and would leave the spirit roaming at large.
In this case, magnetic resonance imaging (MRI) of the brain may well have eliminated any concern that the death was due to shaken baby syndrome. Merrick v. Milledge [21] was a case of an objection not based on any religious or cultural grounds. Mrs. Merrick was a diabetic who suffered regular hypoglycemic shocks. Her husband constantly remained available to assist her on those occasions. However, a circumstance arose where he was forced to be in New Zealand at the same time his wife had to travel back to Sydney. She went to bed and passed away without her husband being able to be available to assist her. The wife had always expressed a view that she never wanted her body interfered with and the husband, for his part, wanted the body to be embalmed and available for viewing. The objection was upheld. Again, virtopsy, perhaps together with toxicology, would have resolved any concerns without cost and trauma of litigation. In New Zealand, there was a very considerable indigenous (Maori) population. The Law Commission of New Zealand
Legal Side
undertook an extensive review of the role of the coroner and the legislation relating thereto. In its report [22], the commissioners discussed the importance of the recognition of the religious and cultural concerns of the Maori population, which also opposed the invasion of the remains of the deceased by the autopsy process. The commission [23] noted the cultural and religious the Jewish and Islamic beliefs, as well as those of Cook Islanders, Fijians, Niueans, Samoans, Tongans, and some Buddhists, in addition to those of the Maori population. The commission noted the concerns of pathologists and other interested parties: [24] The objective of this report is not to espouse an opinion on which perspective is more appropriate, since all views are to be respected. Rather, it seeks to find a balance that meets the interests of the many groups involved, including the deceased, the family and the wider community, while ensuring that the State only intervenes to the minimum extent necessary.
In consequence, the Law Commission recommended and the New Zealand Parliament ultimately adopted in the Coroners Act 2006 what is probably the most sophisticated system of objection so far produced. Some aspects have been noted already. One aspect of the legislation enables the coroner to put aside an objection if urgency demands that the postmortem be undertaken urgently by reason of the loss of evidence should the postmortem not be undertaken quickly or for some other good reason. As commendable as the legislation may be, it is at once obvious that there is still considerable room for dispute. In New York [25], legislation provides that postmortem examinations will not be carried out in the face of religious objection except in limited cases, which generally point to circumstances of possible criminality and matters relating to public health. Clearly enough, the introduction and use of noninvasive techniques will significantly ameliorate such room for disharmony and conflict and, indeed, litigation. The consequence of the introduction of procedures for objection and the intervention of the courts has, of course, given rise to a subsidiary consideration such as the advisement to the family of their rights to objection. In those cases where the postmortem is to be carried out, there are attendant procedures (in the case of New Zealand legislation) enabling the family to attend at the Institute of Forensic Medicine to be close by the body and other steps have been taken to ensure the carrying out of the procedure and other dealings with the body of the deceased in a manner of greater respect than might otherwise be the case. Two instances of cases in which autopsy were avoided should be noted. In the first, an 8-year-old Muslim girl was killed in a motor vehicle accident in December 2004. This was a circumstance in which the coroner historically invariably required an autopsy. Her family objected, citing religious reasons. The body was released to the writer for a CT scan to be performed by a private radiologist. That scan revealed a subarachnoid hemorrhage consistent with the known facts. The coroner accepted this finding, and an autopsy was
© 2009 by Taylor & Francis Group, LLC
33
avoided. The second case involved a late middle-aged Jewish gentleman who was found deceased in his home in which he lived alone. There was no apparent cause of death, no suspicious circumstances, and no relevant medical history. A CT scan was similarly performed and revealed a tumor that impacted on a pulmonary artery. Again, the objection of the family succeeded, and an autopsy was avoided.
A4.1.7 SOME OTHER CONSEQUENCES OF THE USE OF NONINVASIVE TECHNIQUES Societies have to define in what circumstances an autopsy can be undertaken without the prior consent of the relevant party and indeed who that relevant party might be. Further, there has to be guidance as to the extent to which the examination might take place and the fate of organs removed in the process. When the legislature defines the circumstance of compulsive postmortem examinations, it has to consider a variety of competing interests and considerations. The use of noninvasive techniques should and will have a very considerable impact on the redefinition of the circumstances of compulsion. When one refers to the objectives of postmortem examinations, there is a need to consider in respect of each of those objectives when consent is necessary and when, within any particular category of objective or objectives, that need for consent might be overridden—or, alternatively, when the otherwise compulsive force of the law might be legally resisted in a particular case. This makes obvious the fact that any noninvasive process of postmortem examination has the effect of significantly ameliorating and often removing some of the issues described. It is thus possible to avoid not only particular family or community distress but also the need for systems designed to deal with disputation and the cost of maintaining and utilizing the means of dealing with or resolving such disputes. More particularly, the extensive but presently necessary cost of forensic institutes or other appropriate departments of government maintaining counselors and social workers who need to assist families in coming to terms with the autopsy process may be significantly reduced if no bodily interference is necessary. The cost to the community and individuals in dealing with disputes—including those that go to the courts for determination—may similarly be reduced, if not avoided. Accordingly, it is necessary to consider not only the reduced cost per case that flows from the introduction of noninvasive techniques but also other ancillary costs inherent within the present system that may also be reduced. In assessing the value of the information revealed by noninvasive techniques in a particular case, one must take into regard the particular purpose for which the process is to be undertaken or is contemplated to be undertaken. In the case of the autopsy process being used for the sole purpose of scientific inquiry, it would generally be accepted that consent is necessary. In such cases, one would expect
34
consent to be much more readily forthcoming if the process undertaken is noninvasive. There has been a significant complaint within some areas of the medical profession that there has been a significant decline in hospital autopsies [26], which do not fall within the compulsive process of the coroner or equivalent officer. Noninvasive techniques may well lead to a reversal of that trend. This may arise because societies may perceive that it is appropriate to legislate for nonconsensual investigations where there is no interference with the body and no delay in the funeral process, but, to the extent consent might still be necessary, it is far more likely the consent will be forthcoming.
A4.1.8 THE COURT PROCESS The postmortem examination, used for the purpose of the legal process, takes place at several levels. One level is to determine whether a criminal investigation ought to be undertaken. At another level, there is the need to be able to prove or disprove that a particular individual is responsible for the death of another. At yet another level, there is the question of civil liability in compensation that liability may give rise to in a particular case. In any court process, the method by which facts are proved is of critical importance. The transmission of images onto screen in the court, making it capable of showing the whole body and demonstrating the matters at issue, would be a valuable tool available to the tribunal of fact, be it judge or jury or any other body. The tribunal of fact, be it a judge or jury, can more readily appreciate issues with the benefit of such images Alternative scenarios to those propounded by the prosecution may be more readily demonstrated or disproved with such technology. Although second autopsies are performed at the request of defense lawyers, family, or the police, these can be less than satisfactory and sometimes impossible. Photographs taken at the time of autopsy are often sufficient, but in many cases they will in many cases not be as useful as permanently retained images or scans. Defense lawyers can more readily obtain other expert opinions and cross-examine prosecution experts when such images are available. International experts can review images and scans delivered to them electronically and give their evidence by video by referring to the images which all can simultaneously observe. If new evidence comes to light or is sought to be brought to light, images of the entire body and not just photos of those aspects originally thought to be important to photograph can be available.
A4.1.9 CONCLUSION The importance of the use of noninvasive techniques, in particular virtopsy, is immense. The consequences extend far beyond the ability of such techniques to identify particular medical findings. As the technology and knowledge of virtopsy increases, the social and legal consequence will, of course, also increase. Even present technology has had some
© 2009 by Taylor & Francis Group, LLC
The Virtopsy Approach
effect and, if better understood by the courts, coroners, and the police, its benefits would even now be very significant.
A4.1.10 NOTES 1. LL.B(Syd), LL.M(Syd), of the New South Wales Bar. 2. Ian Freckelton and David Ranson, Death Investigation and the Coroner’s Inquest (London: Oxford University Press, 2006), 315. 3. Smith v. Tamworth City Council (1997), 41 NSWLR 680. 4. Coroners Act 1980 (New South Wales [hereafter NSW]), no.27, ss. 24, 48. 5. Anatomy Act 1977 (NSW), no.126, s. 8. 6. Human Tissue Act 1983, no. 164, ss. 7, 8. 7. Coroners Act 1980 (NSW), no. 27, s. 8B. 8. B. Walker, Inquiry into Matters Arising from the PostMortem and Anatomical Examination Practices of the Institute of Forensic Medicine, August 2001, SC. 9. Coroners Act 1980 (NSW), s. 4. 10. Coroners Act 2006 (New Zealand), s. 9. 11. Deitz v. Abernethy (1996), 39 NSWLR 432 (NSW Court of Appeal); at first instance, BC9602510. 12. Van Beelan (1974), 9 SASR 163; Chamberlain v. R (No. 2) (1984), 153 CLR 225. 13. Briginshaw v. Briginshaw (1938), 60 CLR 336. 14. B v. Medical Superintendent of Macquarie Hospital (1987), 10 NSWLR 440; Neat Holdings v. Karajon Holdings Pty Limited (1992), 67 ALJR 170. 15. Stanoevski v. Law Society of N.S.W., BC200510729. 16. (1997) NSWSC 432; BC9902322. 17. P. Vines, Objections to Post-mortem Examination: Multiculturalism, Psychology & Legal Decision-Making, Journal of Law & Medicine, 7 (1999–2000), 422. 18. (1995) 2 VR 176. 19. (1997) NSWSC 432; BC9902322 (see note 16). 20. (1997) 95 A. Crim, WASC, R65: BC9703780. 21. (2002) NSWSC 305; BC200201659. 22. Report 62, August 2000, Law Commission of New Zealand. 23. ibid., p. 62, para. 194. 24. ibid., p. 66, para. 210. 25. New York Public Health Law, s. 4210C. 26. The Royal College of Pathologists of Australia Working Party, The Decline in Hospital Post-mortems: A Safety and Quality Issue for Healthcare in Australia, MIA 180, no. 6, 281–285.
A4.2 AN EXTRAORDINARY DEATH Richard Dirnhofer and Michael J. Thali Positive net results regarding the use of computed tomography (CT)—especially as a supplementary screening procedure— for reporting the findings from external postmortem examinations have also emerged against the legal background of the Swiss postmortem and burial authority laws. This is also against the background of the results of a Japanese investigation [1], in which the importance of employing CT for augmenting the external postmortem examination has been shown (“The usefulness of CT during post-mortem examinations”) and with this leads to further
Legal Side
legal steps for introducing a criminal procedure. “Imaging is a good method for screening and is a useful examination in combination with autopsy.” The Swiss legal regulations may also, thereby, be of international interest. This is because the death certificate is the most important attestation the doctor can give [2]. The discussions and problems of the postmortem examination and the mistakes that have occurred during them are sufficiently well known [3]. This is especially true because the medical and criminal consequences of such mistaken diagnoses appear very quickly and sometimes even drastically. In view of the inadequate legal regulations in Germany one proceeds on the assumption that the number of unreported and undetected murders is just as high as reported ones. The decisive point that we are concerned with is the distinction required on the death certificate between a natural and unnatural death. This raises the question as to whether this provision, which permits only these two alternatives, is able to adequately fulfill its purpose. This is because it contains the assumption that in a postmortem examination the cause of death (e.g., suffocation, fat embolism)—and the type of death derived from it—can be established irreproachably in every case. Everyone knows, though, that the probability of a mistaken judgment must increase if only an external examination of the corpse is made. Coroners and criminologists are aware that the external postmortem examination alone is not an exact examination procedure that is always able to clear up conclusively the question of a natural or unnatural death. The postmortem examination procedure represents much more only a compromise among the requirements of ascertaining the moment of death, the cause of death, the type of death, and the time of death and the necessity, based on economics, of having to perform an autopsy with inadequate means. One must therefore always point out that the kind of diagnostics present in a postmortem examination, due to its nature, only permits a very limited report of the findings and for this reason only allows one to pose a conjectured diagnosis. In normal, everyday cases with believable prehistories, one will have to accept such examinations, and it is also appropriate to do so. With this procedure, though, to want to distinguish a natural death from an unnatural one with certainty places an excessive demand on medical knowledge and ability. Not least, the question as to what is natural and what is unnatural is also not finally and authoritatively solved. In these instances, the Austrian legal system, for example, has opened up the possibility of a supplement through an autopsy, primarily no criminal implications exist. This is possible through the institution of the so-called sanitary police postmortem examination. At the present time, in Switzerland, for such unclear cases—also those that can be declared as such on the death certificate—a two-part postmortem examination procedure is provided before further criminal court proceedings are introduced [4]. Here now is given the explanation of a term hardly known outside Switzerland—“extraordinary death” (Aussergewoehnlicher Todesfall). Its use is of interest, and it has been employed for many decades. The term obviously implies dealing with deaths
© 2009 by Taylor & Francis Group, LLC
35
that depart from the everyday and normal; this distinction emerges from the definition that Fritz Schwarz [5] formulated: “Extraordinary deaths all occur suddenly and unexpectedly, and, as well, are all those deaths caused by violence and where the effects of violence are suspected.” Roughly 10% of all deaths are included in this definition. This definition is based on a phenomen discrimination and keeps clear from the difficult distinction between natural and unnatural death by postmortem examination. This can be elucidated using Figure A4.2.1. The conditions under which the death occurred or a corpse was found exclusively form the decisive criteria for classifying such occurrences into the categories of normal deaths and extraordinary deaths. In an individual case, these characteristics—as they are listed in the left column of Figure A4.2.1—are easy to recognize, both for the doctor called to perform the postmortem examination and for medical laypeople, as they are provided by other persons or by direct perception at the location where the corpse was found or from the corpse itself. In addition, the features of the dying process, or the nature of the discovery, a pleafor-help function pertains which, on such occasions, calls for special attention and caution in the treatment of such deaths. Concerning extraordinary deaths, the highest care and precision in the examination and a critical assessment of the results required. The latter condition is of great importance for the practice, which is why the concept has also found entry into the legal regulations concerning postmortem examinations in Switzerland. The criteria are in accordance with critical criminological thinking. The unknown or dubious prehistory corresponds to the criminological wisdom de omnibus dubidandum and the inclusion of the pattern of clues as well as the suspicious findings on the corpse, a fundamental criminological procedure. The plea-for-help function becomes clear in sudden or unexpected death, not only with heart failures but also with poisonings, electrocutions, child molestations, and when diagnostic or therapeutic incidents during medical treatment take place. For a corpse with (even massive) injury findings, the doctor initially does not have to determine whether the established injuries arose while the person was alive or after death had occurred, such as is seen again and again in corpses taken out of the water or in traffic fatalities. This is because, in these cases, there are sometimes also deaths due to natural causes. The utilization of the concept of extraordinary death does not, though, remove the difficulties of a usable and reliable diagnosis but circumvents it to some degree in order later— strengthened with further information—to permit further differential diagnostic deliberations. The decisive advantage of its use, however, lies in the fact that it makes the coroner conscious from the beginning of his or her task—and the responsibility connected with it—and keeps it clearly before his or her eyes. At the same time, for the practice of the postmortem examination and the judicature the important consequence arises that the mere fact of the presence of an extraordinary death obliges the doctor to notify the bureau of investigation without delay. With this, though, doctors can also absolve
36
The Virtopsy Approach
B. No Medicolegal examination in cases, where:
A. Medicolegal examination in cases, where: Pre-history is unknown or doubtful
Pre-history known and believable
Unusual circumstances surrounding the death, such as in prison, red light district or a discovered corpse
Inconspicuous environment No injuries found on the corpse
Suspicious surroundings such as the pattern of clues The dead person has been identified How death occurred: -suddenly -unexpected -Findings on the corpse that lead one to suspect violence -Corpse with clear injuries Based on phenomenology unclear “so-called Extraordinary death”
Natural death due to - verified disease - normal cause of death Death certificate Notification of the civil registry office
Obligation to notify the police Possible legal implication of the death Examining magistrate’s order for a legal inspection Purpose: Explaining the manner of death Autopsy rate in these cases (30%)
FIGURE A4.2.1 Court systems or coroner’s inquest and procedure in Switzerland. Phenomenological distinctions in a medical examination of the dead person and his or her environment.
themselves of the responsibility in that this is then transferred to respective and responsible examining magistrates. Doctors therefore become sentinels in the critical examination of death. Due to this, the number of unreported cases (e.g., in homicides with obvious judicial implications) can be kept low. In practice, whether the district attorney, then, following such a communication, initiates a criminal investigation is decided only in the next step of the examination, which consists of undertaking a so-called legal inspection. In this examination, the magistrate—together with the public health officer, the local medical health officer, and the police— performs an exam of where the corpse was found; in other words, it is a public health postmortem examination. With this it is guaranteed that a competent judgment is made—not only of the findings and alterations of the corpse itself but also in its surroundings—and that it is performed by those especially trained for such assignments. This regulation also permits a step-by-step approach. This in turn makes it possible to avoid unnecessary expenses. Close cooperation among the examining magistrate, police, and doctor also accelerates the investigation and makes it easier for the coroner to quickly decide for or against opening a criminal procedure. The Swiss legal regulations concerning an extraordinary death—in terms of personnel, material, and finances—thus represents an inexpensive process that in many cases permits a rapid clarification of criminal and civil law questions surrounding a death and, moreover, also satisfies public healthcare requirements. Ever since the Institute for Forensic Medicine at the University of Bern added a computer tomographic examination in cases of these so-called extraordinary deaths, we
© 2009 by Taylor & Francis Group, LLC
find that the results completely confirm what was said in the Japanese publication.
A4.2.1 REFERENCES 1. Hayakawa, M., S. Yamamoto, H. Motani, D. Yajima, Y. Sato, and H. Iwase. 2006. Does imaging technology overcome problems of conventional postmortem examination? Int J Leg Med 120:24–26. 2. Dettling, J. 1951. Legale ärztliche Zeugnisse. In Lehrbuch der gerichtlichen Medizin, ed. J. Dettling, S. Schönberg, and F. Schwarz. Basel: Karger, p. 32. 3. Patscheider, H. 1983. Zur Leichenschau bei außergewöhnlichen Todesfällen. In Fortschritte der Rechtsmedizin, ed. J. Barz, J. Bösche, and G. Schmidt. Berlin: Springer-Verlag, pp. 102–108. 4. Zollinger, U. and T. Plattner. 2005. Der aussergewöhnliche Todesfall–ein besonderer Einsatz im ärztlichen Notfalldienst. Ther Umschau 62:413–418. 5. Schwarz, F. 1962. Grundsätzliches zum außergewöhnlichen Todesfall. Beitr Gerichtl Med 22:298–306.
A4.3 VIRTOPSY AND THE SWISS LEGAL SYSTEM: NEW-EVIDENCE LAW IN FORENSIC MEDICINE? Markus Weber Recent series such as “The Last Witness” (USA) or “Dr. Samantha Ryan” (GB) occasionally give the impression that forensic medicine is able to clear up confusing collections of facts, thereby leading to the conviction of the most clever criminal. Forensic medicine seems to be the king of proof. Against this
Legal Side
impression, though, stands that which is derived from the maxims of modern evidence-based law: Forensic physicians are only then allowed to be active when they are requested by the responsible legal magistrates to answer precise questions posed to them, and ultimately their testimonials are thus still subject to assessment of the evidence. The judge is not thereby bound by the findings; it appears he or she is able to proceed as he or she desires. Is forensic medicine the whore of criminal jurisprudence? For justice to be administered, the task of forensic medicine is to scientifically and understandably document medical findings from both the living and the dead and to analyze and explain these findings. In 2002, Article 252 of the Swiss Penal Code of the Federal Criminal Court (FCC) of Switzerland was added, which regulates examinations on cadavers; in cases of unusual death, if there are indications of an unnatural death—especially if a criminal offense might be involved or if the identity of the corpse is unknown—the article specifies that the Department of Public Prosecution (equivalent to the Office of the District Attorney in the United States) is to order a legal inspection by a medical expert to clarify the type of death or to identify the corpse. If after legal inspection there are no indications of a criminal offense and the identity is known, the corpse is released by the Department of Public Prosecution for burial. Otherwise, it orders the cadaver to be made secure and examined by a forensic medicine institution, even by autopsy if necessary. Characteristically, Article 252 says nothing about virtopsy. The law texts of the preliminary version of the FCC naturally do not exclude the possibility of a virtopsy examination being ordered by the public prosecutor or the examining magistrate following the legal inspection and before the autopsy. Virtopsy is formed from the terms virtual and autopsy; the former is Latin and means something like “virtuous,” while the latter means “seeing with one’s own eyes.” Thus, the Virtopsy Project, developed by the Institute of Forensic Medicine at the University of Bern and by Professors Richard Dirnhofer and Michael Thali, contains a forensic-radiological data bank that corresponds to a virtual “forensic visible man/woman” and produces synthetic model bodies based on this knowledge. This has the following advantages for the future: (1) forensic reconstructive trials and experiments can be omitted; (2) the data is based on computer models; and (3) invasive methods like autopsies will become obsolete. The three main research emphases of virtopsy are as follows: 1. Somatic surface documentation utilizing photogrammetry and optical 3-D surface scanning 2. Physical volume documentation analysis with modern computed tomography (CT) and magnetic resonance (MR) 3. The development of synthetic somatic models for answering reconstructive questions In Article 254, Chapter 5, of the preliminary version of the new Swiss Federal Criminal Co (PV/SFCC), DNA analyses
© 2009 by Taylor & Francis Group, LLC
37
are earmarked in order to clear up a crime or an offense, wherein a sample and a DNA profile are taken as well as measurement examinations from persons convicted of crimes. In contrast to innocent persons, the article also permits examinations of the body and interventions into the somatic integrity, but only when they are imperative in order to clarify a crime according to a certain catalog of offenses (Article 250, par. 4, PV/SFCC). Physical examinations and interventions into somatic integrity are to be performed by a doctor or by another medical specialist. It is deplorable that a virtopsy is not included in the criminal identification process or in physical examinations. Most likely, the principle of free-evidence assessment is so strong that this examination method still has to be tested in order to convince the legislators to elevate it to that level. As a proponent of free-evidence assessment a clever mind like Charles de Secondat, Baron de Montesquieu, would surely today have decided in favor of virtopsy. Since Lombroso the sciences have contributed substantially toward the discovery of truth in a trial. The desire of justice for simple and reliable answers was always larger than the capacity of actually delivering them. It is true that by the 19th century, useful criminalistic methods had been developed. I think, for instance, of the Marsh arsenic sample or the anthropometry of Bertillon. For understanding evidence-based law, every specialist engraves the following sentence in his or her mind: “One sees only what one views and one views only that which one has in mind. Occasionally, though, one has to depend on an arbitrary evaluation of the evidence.” In the second half of the 20th century the picture changed; namely, the clarification of murders has become considerably simpler. The utilization of DNA analyses in crime investigations represents a real breakthrough, and pertinent information can be extracted from such insignificant entities as diatoms. We are approaching a time in which a virtopsy will be able to clear up the most complex crime. It will ill suit a public prosecutor if he wants to keep still and not marvel at the accomplishments of the modern sciences and forensic medicine. Over the decades, forensic medicine developed clinically in that it concerned itself with the living—for example, with the victims of child cruelty or physical injuries. Using forensic autopsies, the dead were examined, and out of natural science developed forensic chemistry and toxicology. A few decades later came forensic molecular biology and DNA research. The analysis methods and the precision of the sensitivity are being constantly refined. In forensic medicine, developments in the microregime and the macroscopic somatic and organ observations followed, and their techniques evolved. In the mid 1990s, imaging procedures like virtopsy from measuring technology and radiology gained entry into the Institute of Forensic Medicine at the University of Bern. The problem as to whether one can establish a three-dimensional relationship between the form of an injury on the body surface and a suspected instrument was the impetus for a research project. For this three-dimensional documentation, methods for anatomic surfaces and objects were evaluated.
38
An additional question was also how the physical interior could be displayed three-dimensionally and noninvasively. In the 1990s optical 3D scanning technologies for the anatomic surface and CT as well as MR for the bodily interior could be identified as being potentially suited imaging technologies. In the last few years these technologies have experienced continuous improvements and refinement. The Virtopsy Research Group correspondingly modified these for forensic and legal medical utilization. Today they are able to record a traffic accident in which vehicles are thrown around just as three-dimensionally as they can the layering of material removed while digging a tunnel that endangered the collapse of an expressway tunnel or the investigation of the running over of a young bicyclist. After the Institute of Forensic Medicine in Bern in close cooperation with the Institute for Diagnostic and Interventional Neuroradiology at University Hospital Bern as well as the Bern Cantonal Police were able to increasingly establish the virtopsy approach for forensic practice, other institutes worldwide have also begun converting to this technology. According to Article 101 of the Bernese Criminal Code (BCC), all the means that science and experience have shown to be valid for discovering the truth are to be utilized. For collecting evidence, bringing in experts is useful (Article 102, para. 1, l. 3, BCC), but only doctors who are personally responsible for the medical opinion are permitted to examine third persons (Article 161, para. 2, BCC), especially when physical examinations or interventions are involved, and the corresponding examinations on the corpse (Article 165, BCC). Regarding measures for collecting evidence, specialists can be brought in and empowered to ask those involved questions (Article 128ff, BCC). As a rule, experts must file a written report with the responsible doctor. This report is binding for the judge and can only be overruled when it is proven to be insufficient or false. Based on the technological developments of the past few years, the Virtopsy Research Project has the goal of evaluating and validating the modern noninvasive three-dimensional documentation and analysis methods for forensic medicine. Trial law is dominated by the written word. Imaging procedures are not excluded, however. According to Article 74, para. 4, PV/FCC the person in charge of litigation can decree that the trial proceedings, in addition to being in written form, can also be registered partially or completely in sound and picture if he or she has informed the parties that are present. The statements of the experts are to be continuously protocolled, and, according to Article 76, para. 6, PV/ FCC, they can be recorded using technical aids. Thus, formal hindrances to the virtopsy imaging procedure do not exist, even when it would have been desirable for the preliminary draft of the FCC to explicitly mention imaging procedures under evidence-collection possibilities. Its advantages are incontrovertible and are threefold:
© 2009 by Taylor & Francis Group, LLC
The Virtopsy Approach
1. It is a noninvasive method for physical examinations of both the living and the dead, meaning that it does not exceed a through examination, as a “legal inspection” and needs no permission from the relatives because the corpse is in no way deformed. Living victims of criminal acts are also not subjected to an invasive examination or treatment (e.g., palpation) but rather only to a surface scan. 2. The imaging procedure is three-dimensional: It makes it possible for the court and public prosecutor to see at once what happened (whether it was an accident or a crime). They are not mere sketches or photographs but rather are moving pictures that clearly portray, for example, the blow, the puncture, the shot, or the somatic impact of the object as well as the results. Reconstructions in foro are possible at any time. Video recordings can be rewound accordingly and can always be commented upon by a specialist before the court and before the litigants. 3. Movement and speed: Reconstructive sketches by the forensic investigator, police photographs of the site of the crime, pictures of the autopsy with bullet and stab canals, and so forth are static. In no case are they able to portray dependably the tempo of a punishable interaction. Virtopsy, however, is able to do this. Virtopsy can demonstrate before a court ad oculos the course of a traffic accident|a car running into a child on a bicycle, for instance|how it manifested at the location where it took place, and how it affected the physical structure of the victim or what traces were left behind on the objects involved. For the future value of the evidence, this is of inestimable quality and quantity. It is to be hoped that the judicature, after the possible crime instruments have been previously documented three-dimensionally (e.g., a ratchet wrench or a shoe sole as the instrument used to deliver the blow), distances itself from the classic forensic documentation using two-dimensional photography and will commit itself to the level of the photogrammetrysupported optical 3D scanner (i.e., to size-relationship-faithful virtopsy). Complex behavior patterns or skin injuries with a specific shape can be matched three-dimensionally to the injury-causing instrument. A possible weapon will be checked via the form, structure, and size of the injury and its form for a possible match. This morphologic “fingerprinting” is now established as a concept. In addition to the aforementioned surface scanning, which permits the documentation of injury shapes, the Institute of Forensic Medicine at the University of Bern also performs magnetic resonance imaging (MRI) examinations on
Legal Side
living persons, for example, with strangulation victims. For instance, a strangulation victim with an injury on the neck with a specific shape can be scanned on the surface, and simultaneously the interior neck findings can be documented with the radiation-free MR tomography. Unlike previously, the interior findings can now be displayed and documented. Based on the injury findings in the critical neck regions, it
© 2009 by Taylor & Francis Group, LLC
39
is possible to place before the court a differentiated opinion regarding mortal danger. All these technical and technological innovations are extraordinarily effective and useful for litigation. Future criminal code must provide space for and anticipate these realities. Only a forward-looking legislation is able to accomplish its purpose of serving the security— before the law—of its citizens.
A5
Religion Stephan A. Bolliger, Michael J. Thali, and Graham P. Segal
CONTENTS A5.1 A5.2 A5.3 A5.4 A5.5
Introduction ..................................................................................................................................................................... 41 Australian Aboriginals ................................................................................................................................................... 42 Buddhism ......................................................................................................................................................................... 43 Islam ................................................................................................................................................................................ 44 Judaism ............................................................................................................................................................................ 45
A5.1 INTRODUCTION Ever since human beings developed a perception of time and space, they started to wonder about their purpose in life, and the notion of the afterlife became omnipresent. This idea of the finiteness of life, or rather the infinity of time, coupled with the desire for purpose and security gave rise to the earliest religious communities. Different cultures and religions arose, with different beliefs regarding the afterlife. One common aspect of most of these early religions was the reverence of ancestors and the deceased in general. The body was often believed to be a link to the immortal soul, or a vessel containing it. Therefore, great care in handling the body of a beloved person was taken. This resulted in different funerary practices, which, depending on the culture, could be enormously complicated, time consuming, and cost intensive. An example for such funerary traditions can be seen in ancient Egypt, where the pharaohs occasionally spent more time and energy on the burial of the deceased than on the needs of the living. In that culture, it was absolutely vital that the body of the deceased was kept intact for the afterlife. For this reason, elaborate mummification techniques were developed. The deliberate preserving of a corpse was, however, by no means unique in ancient Egypt (see also Chapter D8.1 in this volume concerning paleoradiology); mummies are known from almost all continents. Preserving the corpse posed certain problems; the body’s integrity was sacred, thus prohibiting any tampering other than for these purposes. However, certain mummification techniques required the removal of inner organs, for which an incision of the skin was inevitable. In ancient Egypt, the initially necessary incision of the corpse for organ extraction was undertaken by a person who was subsequently outcast and chased away, only to return for the next corpse shortly afterward. Therefore, the necessity of good body preservation, which would have contradicted the religious laws requiring the body being intact, was solved by a rather diplomatic solution. Other ancient cultures had different problems. In the earliest Mediterranean cultures, corpses were often cremated. Although
this meant the most extreme form of destruction of a body, the corpses’ integrity prior to cremation was not to be disturbed. Therefore, the ancient Greeks, who are undoubtedly the masterminds of Western philosophy, science, and medicine, were not able to examine the human body. Their foremost physicians applied the anatomical findings of domestic animals to humans. This shortcoming persisted in Europe into the medieval ages, where persons who performed autopsies were, upon detection, tortured and executed as heretics by the church. The development of religious beliefs and cultures went side by side with the gradual development of judicial systems. Early groups of humans were governed by a family leader, usually the eldest member. However, with increasing population density, interindividual problems grew. Fights with neighboring clans made leaders, who on the other hand also had the possibility of suppressing weaker individual groups, necessary. These leaders wielded the power of punishing or rewarding certain individuals within their own group or clan. Probably around this time, a notion of justice arose. In the beginning, the subjective verdict of the leader, or preform of judge, was sufficient to convict a person. However, ancient philosophers and scientists gradually challenged the established, absolutistic social views. With growing knowledge and increasingly differentiated views on life, the population required more proof for the guilt or innocence of an individual. This claim is best demonstrated by the Roman law in which is stated, In dubio pro reo, meaning “in the case of uncertainty,” judge in favor of the defendant. This pressure on judges made the acquisition of findings, or rather “hard facts,” necessary. Concerning human corpses believed to be victims of a crime, this meant also gathering findings both on and in the corpse. (For an overview of the history of this collection of internal findings, also referred to as autopsy, see chapter A3 in this volume, section on autopsy.) The three pillars of human society—religion, science, and law—were, and still are not, necessarily compatible. Although law and religion often go hand in hand, as the Sharia in Islam clearly demonstrates, the ordering of certain 41
© 2009 by Taylor & Francis Group, LLC
42
examinations—especially autopsy—may cause friction with religious or cultural rules. Science, on the other hand, regularly clashes with religion. Although the days of persecution of scientists, as witnessed in medieval Europe, are mostly over, the social rift between religion and science has not been overcome. Indeed, the current discussion between evolutionists on the scientific side and creationists on the religious side, for example, seems to be intensifying. The great challenge of a modern judicial system is therefore to provide solid findings while also respecting religious laws and cultural customs. To date, the examination of deceased persons often requires an autopsy, which as mentioned before means tampering with the body, an act rarely welcomed by religious groups. A selection of the opinions on autopsy and the problems encountered with this procedure was presented by different religious groups at the Postmortem Radiology Conference in Sydney, Australia, in 2005. Of these, transcripts presented by representatives of four religions—native Australians, Buddhism, Islam, and Judaism—are cited herein. These are cited in alphabetical order and by no means imply the authors’ preference of certain religions.
A5.2 AUSTRALIAN ABORIGINALS Talk presented by Ray Jackson, head of the Indigenous Social Justice Association. (Quoted with permission) What I’d like to do is to present some understanding of the burial rituals of our people around the country and link that in to the fact that autopsies were unknown, body mutilations and things like that were not known, and we now mainly fall under the Christian burial system although there is a fairly large group who are joining the Muslim faith, and I would imagine that there would be some black fellas out there who are Buddhists or Jews or whatever faith there is. I thank Richard for the blessing of the conference, and I recognize again that we are on the Stolen Lands of the Cadigal Clans of the Eora tribes within the wider Dharruk Nation. I thank the NSW Joint Committee on Post Mortem Examinations, of which I am a member, for inviting me to speak here today. I especially thank Graham for the work that he and others have put in to getting this conference up; I think it is a most needed conference, especially when you look at the autopsy practices that are still going on today. To start with I emphasize that autopsy, or body-invading mutilation, was not practiced by the Aborigines ever. It was not part of the scene. Even your enemies were respected, and they were not mutilated. The Royal Commission Recommendations arising from the Deaths in Custody Inquiry called for pathologists to forego autopsies and for the Elders of the Families of those who died in custody to stop the bodies from being touched. The pathologists ignored this, and the practice continued. Recommendation 25 of those recommendations allowed for a medical person picked by the family and also an Aboriginal person picked by the family to attend the autopsies: one for scrutiny of the medical practices
© 2009 by Taylor & Francis Group, LLC
The Virtopsy Approach
used; and my role (as it was me) fell to making sure that the cultural respects were shown to the deceased. And I can happily say that with all the autopsies that I did attend of my people, those cultural respects were shown by the pathologists. Traditionally there are over 500 language groups pre the 1788 invasion. There would have been roughly the same number of beliefs and practices. Prior to looking at the burying rituals and practices it first becomes necessary to understand (or at least to try) the thinking of the society that brought those beliefs and practices about. I quote from Australian Dreaming: 40,000 Years of Aboriginal History (Lansdowne Press, 1980). Now I quibble with the figure 40,000. We believe it’s at least 160,000 years that our culture has been on this land. In any society the death of an individual comes as a threat to the whole fabric of life. Among Aborigines, however, death is seen as inevitable only in the very old. The death of a healthy man or woman must have been caused by the evil magic and sorcery of some enemy of the dead person. Natural causes of death are seldom admitted except in the case of the very young, babies, and old people, all of whose deaths cause little disruption to the pattern of society. But when an adolescent or a person in the prime of life dies, there is a sense of great loss, and the whole camp goes into deep mourning. We now call that mourning “Sorry Business.” Once sorcery is suspected, an inquest must be held to determine the identity of the sorcerer and exact retribution from him or his relations. A common method of finding the sorcerer is to look at the cleared ground of the grave; a hole in the smooth dust could indicate the direction of the country of the murderer—or even insect tracks could reveal the whereabouts of that murderer. A careful consideration of all the facts is given by the fully initiated men of the tribes. Sometimes a corpse itself is interrogated. As questions are asked about possible murderers, the spirit of the dead man is believed to cause the corpse to jerk around at the correct name. Once the murderer is established, a settlement is reached with the dead man’s relations to determine the price to be paid or the punishment to be exacted. Even with the advent of Western medicine, although the European explanation of death might superficially be accepted, the people still look beyond that and attempt to discover who caused the death especially for the more traditional people. In Carnarvon, Western Australia, a man who was drunk fell off the back of a truck and was killed. Two years previously this man had been involved in a fight with another man over a woman. The other man took a rifle and while trying to shoot the first man accidentally shot and killed his own uncle (his mother’s brother). The second man was convicted of manslaughter and jailed, but the Aborigines regarded the first man as being to blame for the death. This man was told he would have to go through the law and be subject to certain tribal punishments, but he successfully evaded this. When it was learned that he had fallen off a truck, some of the older men declared that he had finally been punished by the spirit of the dead uncle that had lain in wait for an opportunity to take his life and had pushed him off the truck.
Religion
Burials can occur by placing the body in a grave, in a tree fork, in a cave or wrapped in bark, or in such places that would be conducive for rest and peace for the dead and also for the living. As David Mahwaljari explains: Man’s spirit exists before birth in its totemic birth place and this spirit will leave him and continue to exist after his death. When a man dies his true spirit must therefore be hastened to its proper resting place, the place of the totemic origin, which can also be interpreted to mean the Dreamtime. If the spirit remains around the camp it will harass and frighten his relatives and therefore all precautions are taken to keep this wayward spirit out in the bush with the other spirits of the dead. In Northern Australia and even as far south as Northern NSW the patches of rainforest jungle that sporadically occur are much feared and avoided as haunts of the spirits.
Even today in Sydney’s coastal caves there are many Aborigines who will not venture near them due to the spirits that they contain. On Bathurst Island north of Darwin such rainforests are only entered by many people together and only at the brightest time of day. In order to diminish the attraction of the familiar surroundings to the dead man, his body may be smoked with burning leaves and the limbs broken and tied to prevent his spirit from reanimating his body. In some areas the widow or the mother of the deceased may become emotionally dumb for several years. The tracks of the dead person, both physically and metaphorically, are swept away, and the camp where he died or last lived is abandoned and avoided. His name is never mentioned, and even words with a similar sound are not used. Funeral ceremonies are very protracted and in the Northern Territory can go on for years. The Tiwi Aborigines on Bathurst Island and Melville Island use the intricately carved burial poles. In Arnhem Land they place the body in hollow logs that are specially made. In Queensland the body is wrapped in painted bark. Prior to his death his relatives gather around and sing to him instructions on what to do when he is dead. Upon his death the relatives then cut themselves to show their grief. The mortuary ceremonies are completed in stages over many months and years, and then the bones are delivered into the custody of the man’s mother, who then keeps them until the appropriate hollow log is ceremonially made. Sydney’s burial practices included food restrictions, which were enforced for some family members, and the name of the deceased not being used again. Living people with the same name will change their name to another. The two most common burial protocols were cremation followed by burial or just burial. A shallow grave is formed and lined with grass, and the grave is filled in. For cremation a deeper grave is dug and lined with grass and wood. Grass is then spread over that, and the body placed is upon it on its side. Any grave goods such as spears or baskets are placed around the body, and the body is covered with logs and grass and then lit. The next day the bones and ashes are gathered and covered with bark and then covered with logs. Throughout the Dharruk Nations the practices varied as to what part of the Dharruk
© 2009 by Taylor & Francis Group, LLC
43
Nation the person died in, whether it was the coastal, hinterland, or mountain areas of the nation. The Sydney Aborigines mostly now have a Christian service and burials or cremations. The one factor that remains solid throughout the history is the horror due to the body being mutilated or body parts becoming missing, because great respect is always shown to the dead and because autopsies are seen as being like a double murder. Autopsy practices have always been at best grudgingly accepted, and the reasons that may require an autopsy go by without real understanding. I have attended several autopsies authorized by my role as the Management Committee Co-ordinator of the Watch Committee, and I’ve always had difficulties attempting to inform families why body samples are taken. It doesn’t matter whether those samples are taken for what is alleged to be scientific knowledge; families want their loved ones whole. They don’t want anything taken from them. I was not willing to tell them that their loved one’s brain was generally taken for further tests. The families would never have accepted that and still don’t. During the infamous death-in-custody era from 1980 until the present day, there were just too many cases where the body was handed back with body parts missing—sometimes internal organs, a heart, a liver, other things were taken for further study and were then disposed of later. As I said, the brain, as you would know, was always taken—missing bones, all that sort of stuff, but always the brain was taken. Discussion at times with Professor John Hilton during the mid 1990s led me to no real understanding of why there is a need to take the whole brain rather than merely a sample, and I still don’t understand why it is necessary to take the whole brain. Our customs and rituals have never included mutilation of the body, but with autopsies to date this is what happens. The processes being pushed by this conference are, I believe, the only real humane way to treat the deceased. There will be a time when full or part autopsies may be required, but that is not always the case. When it became public knowledge among the communities that bodies had been buried or cremated without the brain, there has been much pain and suffering—a terrible reliving of the Sorry Business—and I have had to deal with several families because I was there at that autopsy and they wanted to know why I did not tell them that the brain was taken. I still don’t have a good answer. Thank you for your attention, and I trust that this conference will successfully push for a better way in the treatment of the dead.
A5.3 BUDDHISM Talk presented by Jack Heath, president of the Sakya Tharpa Ling Tibetan Buddhist Institute of Meditation in Sydney. (Quoted with permission) The first reason is that Tibetan Buddhism holds very sophisticated views on the process of death and the way in which consciousness leaves the body and subsequently takes rebirth. And because reincarnation is a core belief of
44
Buddhists, the issues being addressed in this conference bring added importance that might not apply in the context of those who don’t believe in reincarnation. There is in fact a very famous text called The Tibetan Book of the Dead, aspects of which were popularized in the wellknown book by Sogyal Rinpoche The Tibetan Book of Living and Dying. For many Westerners this has been their introduction to Tibetan Buddhism. I don’t propose to go into these views on the death process other than to say that how the body is treated after death, including the environment in which it is placed, is of profound significance in Tibetan Buddhism. Tibetan Buddhist custom is that the body should not be touched or disturbed for three days after death—for autopsies or cremations. This is particularly important in the case of advanced spiritual practitioners, and it is believed that if the body is touched in a certain place—for example, if an injection were given—that it may draw the consciousness to that spot and cause the consciousness to leave the body in a way that leads to an unfortunate rebirth and increased suffering in a subsequent life. There is one famous Tibetan Master who is reported to have told people who were complaining that the corpse might smell if it was kept for three days, particularly in hot weather; he said, “It’s not as though you have to eat it or try to sell it!” So strictly speaking, autopsies and cremations are best done after three days. It is, however, possible that if special practices are performed and certain results are achieved, this might occur earlier. The second reason why I appreciate the opportunity to speak today is that I mentioned the importance of not touching or disturbing the bodies of advanced spiritual practitioners. This is of particular significance to our center, which was established here in Sydney in the late 1980s by Gyalsay Tulku Rinpoche, who is recognized as the 14th reincarnation of a long line of Tibetan Masters stretching back many centuries. In many ways he was the father of Tibetan Buddhism in Australia in that many of his current practitioners from the four schools of Tibetan Buddhism—certainly throughout Sydney—took their first teachings from him. In September 1993 at a relatively young age Rinpoche passed away unexpectedly while visiting Canberra. In the immediate aftermath, this resulted in significant conflict between the students of Rinpoche and the authorities in Canberra who wanted to perform an autopsy immediately. It was from this event and over the subsequent years that there emerged a number of legislative and regulatory changes that apply today and that are designed to take into account the spiritual beliefs of people who have passed away. The third reason why it is important for me to have this opportunity here today is to stress the importance of the interfaith dialogues, and this in fact is one of the key themes that His Holiness the Dalai Lama talks about on many occasions. So I very much appreciate the different spiritual traditions given the opportunity of voicing our views concerning this forum. And finally, the fourth reason why I value this opportunity concerns the important interaction between Buddhism and science. His Holiness the Dalai Lama has always said
© 2009 by Taylor & Francis Group, LLC
The Virtopsy Approach
that if science can prove that a particular Buddhist view or belief is not correct, then Buddhists are bound to reject that belief. In those areas, however, where science is unable to be definitive, then His Holiness asks that we keep an open mind until proven otherwise. I think this is particularly important around issues such as postmortems, taking into account the many centuries of learning, inquiry, and analysis that have been undertaken on the death process in Tibetan Buddhism. I note that there are considerable discussions in this conference on the use of magnetic resonance imaging (MRI). Some of you may be aware that over recent years His Holiness the Dalai Lama has been involved in what is referred to as the “Mind and Life Dialogues,” which involve a number of leading international neuroscientists and other scientists and psychologists. These have taken place in Dharamsala in India and Harvard University and are touched upon in Daniel Goldman’s book Destructive Emotions, which some of you may be familiar with. This book discusses a number of experiments undertaken with monks using electroencephalography (EEG) and functional MRI, which involves not just photos but also video recordings of brain activity including showing how the brain performs in states of deep meditation. There is one interesting example where a monk in deep meditation is virtually unmoved when a gunshot is fired in immediate proximity to his head. I’m not sure about the scope for using functional MRI postdeath and also whether or not that would be appropriate in the Buddhist tradition. But I’d like to just finish and leave it here and say thank you again for the opportunity to contribute to this forum. I hope that your deliberations lead to great insight and understanding that enables us to take into account the various belief systems around the process of death and how the body should be treated postmortem. At the end of the day, though, I have no doubt that we all share a very strong common desire to undertake this research in a way that leads to longer and happier lives for all of us.
A5.4 ISLAM Talk presented by Ameer Ali, chairman of the Australian Federation of Islamic Councils (AFIC). (Quoted with permission) When a Muslim dies, the body is wrapped in a shroud, the color of which material should preferably be white because the Prophet was covered in white. In some cases, depending on the country of origin, colored shrouds are also permitted. The fabric should not be expensive but of good quality. The cost should be paid by the family of the deceased. If the deceased has no family, then the cost should be borne by the community. The men are wrapped with three pieces of clothing: a long, sleeveless, ankle-length shirt; a lower sheet on which the body is placed covering from head to toe; and a longer sheet placed over the body also covering the body from head to toe. Both sheets can be tied together. Women are dressed in the same way, with an additional sheet to cover their head
Religion
and face and another to cover the body between the breasts and thighs. The funeral prayer is very important. The purpose of the prayer is to ask God to forgive the sins of the deceased. A group of people should attend the funeral prayer and pray together. The prayer is mandatory. Generally speaking, the funeral prayer need not take place inside the mosque (although it is done so in Australia) because a mosque is a place meant for the living and not for the dead. So the prayers are normally said in a room in which the body lies or in the cemetery in some countries, and the Imam or the one who leads the prayer must face Mecca with the body laid perpendicularly in front of him. And also—though this may not be relevant to you people—normally in the Muslim prayer there is a series of prostrations, but for the funeral prayer there is no prostration and all the bending and bowing does not take place. If the deceased committed suicide, in some countries the Imam is not allowed to say the prayer but a member of the group or a Muslim volunteer can say the prayer. In all cases, the body must be buried as soon as possible, preferably within 24 hours. If there is a case for medical or legal investigation then some jurists have said that there may be some delay, but before the body decomposes it must be buried. What about autopsy? The Koran, the primary source of laws in Islam, is silent on this. It neither prohibits nor sanctions autopsy. This may be understandable because autopsy is a very recent development and the need for autopsy arises more from the demands of law and criminal procedures rather than from the demands of medical science. So in the history of Islam, when jurisprudence and medical science developed hand in hand, the Islamic legal system did not have a place for autopsy. However, there were religious sanctions against autopsy at that time. That is why in the literature on jurisprudence we don’t find much about autopsy and whether it is allowed or not. However, even though the Koran is silent, there is a hadith or a saying of the Prophet that refers to human bones specifically. It says, “Do not break it, for breaking this bone even after the death of its owner is like breaking it while he or she is still alive.” This hadith is the basis on which people say autopsies are not allowed in Islam because you are disfiguring the body. However, there are two principles in Islam that help overcome this prohibition. One of them is the principle of darurah, which means necessity. Necessity makes what is prohibited lawful. I’ll give an analogy for this. You all know that Muslims, like the Jews, don’t eat the flesh of swine, but if that is the only food that is available to you to survive in a desperate circumstance then the principle of necessity allows you to consume that. This is one instance where even the Koran makes that which is prohibited permissible. Necessity dictates exceptions. What is prohibited can be permissible under that principle. The second principle is falah, which is generally translated as “welfare” or “common good.” If the general welfare of humanity is enhanced by a certain action then that action, according to this principle, should be encouraged. This is where the relevance of autopsy comes
© 2009 by Taylor & Francis Group, LLC
45
in. For reasons of improving medical knowledge that benefits humankind, autopsy becomes permissible in accordance with this principle. The relevance of necessity to autopsy can be explained with an example. Suppose a person is charged by the court of law for murder, and suppose the fate of the accused depends absolutely on the evidence of an autopsy. Islam, according to this principle, does not stand in the way to stop the autopsy procedure. Of course, as a matter of courtesy, permission must be obtained from members of the family of the deceased to carry out the autopsy. Even that may not be absolutely necessary in all cases. However, the body must be treated with utmost respect, and even in the case of removing organs only the organs that need to be examined should be removed and must be placed back in the body after examination and before burial. These are some of the principles that govern autopsy in Islam. To conclude, the Koran is silent on the matter. However, a saying of Prophet Muhammad is interpreted as a prohibition of autopsy. Yet the principles of necessity and welfare that govern all actions of Muslims provide room for autopsy procedures. Many of the current rulings on autopsy in the Muslim world are based on these two principles. One should not forget the fact that legal procedures in Muslim countries are still at a developmental stage. Many Muslim countries are practicing a mixture of Shariah laws and Western laws. There is a continuous debate going on between the religious jurists and experts in other disciplines on several medical and legal practices and issues. This is a healthy development. The debate is young in Islamic countries, and therefore you can expect controversies in different parts of the Muslim world on this subject. But as far as my knowledge goes, and from the advice I was given by Imams, the conclusion is that the principles of necessity and welfare permit autopsy and even organ transplants.
A5.5 JUDAISM Talk presented by Rabbi Moshe D. Gutnick, Sydney Beth Din. (Quoted with permission) Jewish law is a complex structure based upon original Biblical injunction followed by case law comprising judgments that were made by the High Court of Israel (known as the Sanhedrin, or “group of 70”). These rulings were incorporated in the Talmud some 1,500 years ago and form the basis of all Jewish legal decisions. The Talmud was followed by 1,500 years of case law, individual response, and codification. It is a legal system as thorough and complex as any other, and it is based upon the assessments and judgments of the various judges who have made legal decisions throughout the centuries. The difference, however, between the Jewish legal system and other secular legal systems is that the Jewish system is predicated upon a belief in certain basic fundamental principles that we believe are God given. They revolve around the understanding of various verses in the Bible and the transmission of what is called the Oral Law, which was given to
46
The Virtopsy Approach
Moses at Mount Sinai and expressed in all the judgments of the High Court of Israel. The practice of forensic medicine is indeed studied and ruled upon in the Talmud, which was redacted into its current form between the years 300 to 400. The Talmud records a discussion in relation to a person who had been put to death or had been sentenced to death for committing the capital crime of murder. Jewish Law mandates that murder is a capital offence and liable to the death penalty if indeed the person whom he killed was a healthy individual. But if the person who was killed was someone who was going to die anyway from injury or terminal illness, then while the crime is considered murder, the perpetrator—even if found guilty—was not put to death. Put simply, if God forbid a person was in late-stage cancer and someone murders that person, the murderer is guilty of murder but could not be put to death for the capital crime. The Talmud then discusses whether one is required to perform an autopsy on any murder victim. If as we have just stated, a person is already going to die from another illness anyway, then the murderer cannot be put to death, so we should be required to perform an autopsy on any victim to find out as to whether the cause of death was indeed the murderer or whether there was some other cause involved in the actual death or if the person was terminally ill. The Talmud comes to the conclusion that such an autopsy is not permissible because there is a prohibition of mutilating the dead. The Talmud then asks that since we have an overriding rule that all prohibitions are suspended for the sake of saving life we should be able to perform an autopsy in order to save the life of the murderer. The Talmud concludes that because the autopsy would not be conclusive, the performance of the autopsy remains prohibited. We see from this passage of the Talmud two principles: 1. There is a prohibition against mutilation of the dead. 2. (Just as important if not more important) The prohibition would be suspended if, as a result of the autopsy, one could save a life. The only reason why the autopsy was not permitted in this case was because the results could never be conclusive, but if in practice forensic medicine could be developed to a level where there were conclusive results, then in order to save a life autopsy would be permissible. There is another example of autopsy in the Talmud. As in most legal systems, the actions of a minor have no validity in law. The Talmud records an instance where a particular person, a young man, sold properties of the family and then the young man died. The rest of the family, those who would normally have inherited those properties had they not been sold, argued that the young man who had sold the properties was indeed only a minor and therefore the sale was invalid and they were entitled as part of their inheritance to get those properties returned to them from the purchasers. They demanded an exhumation and autopsy
© 2009 by Taylor & Francis Group, LLC
to prove the young man’s age. Rabbi Akiva, a famous sage, forbade the autopsy and again for the same two reasons. The first reason again was that you are not allowed to mutilate the dead. The second reason given is that because once the body was placed into the grave there was again no way of conclusively being able to prove the age of the deceased. Therefore, he forbade the autopsy. We see from these cases that already some 2,000 years ago, legal discussions in relation to autopsy were taking place and specifically in terms of forensic medicine—both in a criminal case and even in relation to a civil case. Today Jewish law has basically not deviated from the aforementioned principles. Autopsy generally is not permissible; in fact, any ill treatment of the dead is not permissible, and the cadaver must be treated with dignity and respect. This is for two main reasons. The first is out of respect for the living, as the cadaver was the vessel that contained in it an eternal soul, a spark of life—we sometimes use the terminology “a spark of Godliness”—so there is holiness attached to that body. It is not just an empty vessel of dead tissue but rather the vessel that once contained a spark of divinity that now retains some of that holiness even after the soul’s departure. The second is that the Jewish lore teaches that until the body is buried the departed soul still has some attachment to the body and therefore feels in some way the pain of the body. So if one mistreats or cuts or causes any damage to a body, that causes pain to the soul. Similarly, even if you delay the burial of the body, the soul is looking to “rest in peace” to find its way to Heaven and cannot do so until the body is buried. The soul isn’t allowed its freedom until the body is buried. There is a specific biblical verse that commands immediate burial of the dead, as well as the mystical verses “and man shall return to the dust” and only then “and the spirit returns to the Lord that gave it.” However, all laws are suspended when life is in danger. The Bible says, “I have given you the commandments to live by them” and, God forbid, not to die by them—every commandment is suspended for the sake of life. Therefore, not withstanding all this Jewish law states that in a situation where there could be direct immediate assistance to save a life, then all the laws are to be suspended and an autopsy is permissible. An example of this would be in the case of a plague that is affecting a community—an epidemic or even a pandemic—and in order to be able to find a cure to save lives, it would be permissible to perform an autopsy. Similarly, in order to catch a murderer—who will murder again unless caught—it would be permissible to perform an autopsy. However, there is a qualification that the saving of life that suspends prohibitions to be immediately apparent, “something which is directly before you.” Therefore, taking a sample so that sometime down the road we may do some experiments on the various limbs or organs, or just because we may think that there may be some potential benefit from performing an autopsy, under those circumstances Jewish law doesn’t permit an autopsy because the life-saving benefit is not immediate. Of course, there is a great deal of discussion as to what does immediate mean,
Religion
as it can be a very subjective term. One of the interesting examples is in relation to the understanding of a genetic illness. Let’s say there is a particular genetic illness that involves a particular family and someone dies as a result of that genetic illness. Are you allowed to perform an autopsy on that cadaver in order to find a cure for that genetic illness? Now at the moment there’s nobody dying from that illness, but because it is a genetic illness it is clear and apparent that that illness will again appear in the next generation and the generation after that. Most authorities consider the fact that the existence of the genetic illness is immediate within all the family, even though the family is not at the moment suffering symptoms from that illness, an autopsy would be permissible. There are various applications of these rules, and they require case-by-case judgment by a qualified rabbi; however, in general terms a cause that will in a direct way lead to the saving of life would be sufficient to allow autopsy.
© 2009 by Taylor & Francis Group, LLC
47
It is important to note that even when an autopsy is permitted the most minimally invasive methodology is to be used. This, again, is to protect the dignity of the deceased and to prevent that notion of pain that exists as a result of the soul still being connected to the body prior to burial. In that regard the use of radiology, MRI, and other noninvasive methods is very strongly encouraged by the Jewish community because that is a way of satisfying both the need to find the appropriate cause of death while maximizing the respect for the dead. Science is capable of bestowing upon humanity the greatest blessings of life and happiness and God forbid the opposite. If we as a community enshrine in our laws respect for both the living and the dead—and find resources with which to satisfy both our scientific needs while protecting always the sanctity of the living and even the dead—then the science discovered will always be a blessing. If we fail to protect human dignity, then the science isn’t worth having.
Part B Imaging and Visualization Methods/Explanation of Techniques
© 2009 by Taylor & Francis Group, LLC
B1
External Body Documentation
Ursula Buck, Silvio Näther, Marcel Braun, and Michael J. Thali CONTENTS B1.1
Historical Development of Geometric Documentation of Wounds in Forensic Medicine .............................................. 51 B1.1.1 Measurements by Hand ...................................................................................................................................... 51 B1.1.2 Photographic Documentation ............................................................................................................................. 51 B1.1.3 Photogrammetry ................................................................................................................................................. 52 B1.1.4 3D Surface Digitizing (Scanning) ...................................................................................................................... 55 B1.2 Photogrammetrical 3D Documentation ........................................................................................................................... 56 B1.3 Optical 3D Surface Scanning .......................................................................................................................................... 58 B1.4 Photogrammetry and 3D Surface Scanning of the Deceased ......................................................................................... 59 B1.5 Digitizing of Injuries of Living Persons .......................................................................................................................... 59 References ................................................................................................................................................................................... 60
B1.1 HISTORICAL DEVELOPMENT OF GEOMETRIC DOCUMENTATION OF WOUNDS IN FORENSIC MEDICINE B1.1.1 MEASUREMENTS BY HAND Since initiation of photography for documentation of accidents, investigating authorities have asked the police for geometrical dimensions. These dimensions should compensate for the loss of information caused by distortion in the photographs. First of all, the accumulation of traffic accidents led to attaching great importance to documentation of traces. Answering questions regarding dimension of skid marks and positions of the cars in an accident plays an important role during the preliminary proceedings. A photograph from 1921 (Figure B1.1.1) shows that in those days measurements by hand were done and drawn into the photograph and, accordingly, into plans. Even today, measurements by hand are suitable means to collect the most important metrics when other instruments are not available.
B1.1.2 PHOTOGRAPHIC DOCUMENTATION In 1838 the French painter and physicist Louis Jacques Mande Daguerre invented a process called Daguerreotyping to fix latent images. This process was the foundation of contemporary photography. Daguerre exposed iodinated silver panels at the back of a camera obscura. He used “short” shutter times from 3 to 30 minutes, depending on season, time of day, and weather. Then he developed the panels by using quicksilver. The Daguerreotypes soon became very popular, and his invention was distributed under license by Alphonse Giroux Paris.
Talbot invented the negative-positive process in 1839. With this method it was possible to copy any number of paper-positive patterns from one negative, in contrast to Daguerreotypes, which were unique. In 1840 Petzval calculated objectives with a luminosity of 1:3:2, which was fantastic for those days. Thus, it was possible to shorten the exposure times conspicuously whereat the depth of sharpness was extraordinarily high. The first camera with brazen body and this highly luminous objective was produced and brought to market by Voigtländer from Vienna. With improved developer solutions and developing techniques, exposure times in the range of seconds could be reached. After that photography entered the areas of documentation and reproduction. Since 1900, photography has been more and more common in forensic documentation. For nearly 100 years photography has been used to capture visible and noticeable findings during the necropsy [1]. In 1912 Kratter mentioned photography in connection with forensic questions for the first time, and today it is a fundamental part of the documentation. Overview photos of bodies and detail views of injuries serve as mnemonic devices, visual aid, and visual exhibit. Photography is based on the principle of central projection. In contrast to a parallel projection, the scale is not unique over the whole image. A yardstick placed in the object space provides an approximate proportion of the object. But because of perspective distortion it is only correct in the layer of the yardstick. The perspective distortion can be minimized by taking the pictures with large focal length and orthogonally to the object, and an angled scale should be placed in the object layer. However, this method is not effectual for using pictures for detailed reconstructions and for taking accurate measurements. 51
© 2009 by Taylor & Francis Group, LLC
52
The Virtopsy Approach
FIGURE B1.1.1 Historic photograph of a traffic accident in Zürich (1921). Dimensions and additional information for the analysis were drawn in the picture.
A three-dimensional object is documented with twodimensional pictures. For accurate depiction of the proportions of an object, photogrammetry must be performed.
B1.1.3 PHOTOGRAMMETRY In 1933 stereophotogrammetry was introduced to the Swiss police. In Germany this method has been used since 1950, predominantly for documentation of traffic accidents. Later it was also performed in cases of capital crime (Figure B1.1.2). In Germany in 1987, Baden-Wuerttemberg was the first state to replace stereophotogrammetry with multi-image photogrammetry [2]. In Switzerland, multi-image photogrammetry has been in use since 1989. For 3D documentation in forensic and traffic accident documentation, this method is the leading technology because of its easy and flexible application. An important advantage of multi-image photogrammetry is that even pictures captured at the scene with commercially available cameras can be included in the analysis. For the preservation of evidence in forensic medicine, photogrammetry was used for the first time in 1996 at the Institute of Forensic Medicine in Bern in collaboration with the Scientific Forensic Service of the Zurich City Police. The photogrammetric three-dimensional documentation of patterned injuries of the skin and bones in forensic medicine enables a computer-aided analysis of the injury and the visual correlation of the wound and inflicting tool. The temporal independence and the preservation of the facts of the
© 2009 by Taylor & Francis Group, LLC
case, which allows for analyses at any time, even after years, are big advantages of the method. The following murder case was the beginning of the application of photogrammetry in forensic medicine for the 3D documentation of the morphology of wounds. In the region of Bern, some years after a murder the question occurred as to whether a wheel wrench out of a car was the tool that had caused injury to the victim’s head (Figure B1.1.3). An identical tool was purchased for comparison. Photographs of the injury to the head and the cranial bone were analyzed by the method of photogrammetry. Thus, it was possible to find correlations between the supposed injury-causing tool and the injuries (Figure B1.1.4 and Figure B1.1.5). With means of photogrammetry, three-dimensional coordinates of discrete points of an object can be determined by measurements made in three or more photographic images taken from different positions (Figure B1.1.6). Before photographing, reference targets need to be stuck or painted on and around the object, and one precise measurement of distance (i.e., reference line) within the area of the object has to be established as a known. Additional “natural marks” (e.g., dotsized blood stains or skin impurities) can be used as reference targets (points). The higher the number of points acquired, the more detailed the object can be recorded and later represented. To increase the number of identifiable points, a grid of small crosses is projected onto the object’s surface with a slide projector (Figure B1.1.7). In addition to the photograph of the wounds without projection, another series of
External Body Documentation
53
FIGURE B1.1.2 Photogrammetry equipment of the Police of Bern, Switzerland, in the 1960s.
FIGURE B1.1.3 In a murder case, the question arose as to whether a hammer or a wheel wrench out of a car caused these patterned injuries.
© 2009 by Taylor & Francis Group, LLC
54
The Virtopsy Approach
FIGURE B1.1.4 As result of the photogrammetric analysis of three images of an injury, the injury lines could be drawn in computer-aided design (CAD).
photographs of the object with the projected crosses is taken and photogrammetrically analyzed. Marcel Braun introduced this procedure to the Institute of Forensic Medicine at the University of Bern in 1996 [3,4]. This method is easy to use with inexpensive equipment but is time consuming, because to calculate the 3D coordinates of the surface points all points have to be measured by hand in at least three images. The point spacing of the
grid varies from 2 to 5 millimeters, depending on projection distance and curvature of the object. Besides many advantages of photogrammetry, the limited number of acquirable surface points, the length of time needed for analysis, and therefore the limitation to a few relevant areas of the object are disadvantages. The grid projection method was a first step in the direction of 3D surface scanning.
FIGURE B1.1.5 The bone injury was compared with the presumable injury-causing instrument. This was the first time that a computerbased 3D analysis was performed and is the first step into the future of 3D data-based forensic reconstruction.
© 2009 by Taylor & Francis Group, LLC
External Body Documentation
55
FIGURE B1.1.6 Images series for the photogrammetic documentation of skin injuries (simulation with a gypsum model). At least 8 image points have to be measured manually in each image. Therefore, crosses are drawn on the object in the area of interest, and a solid scaled ruler is placed for the definition of a reference line.
B1.1.4 3D SURFACE DIGITIZING (SCANNING) In 1998 Subke introduced the 3D surface scanning method for the forensic documentation of the body surface [5,6]. The Centre of Forensic Imaging and Virtopsy at the Institute of Forensic Medicine of the University of Bern
(IRM, www.virtopsy.com) and the Scientific Forensic Service of the Zurich Municipal Police use the GOM ATOS XL digitizing system for their work. This system can easily and precisely measure smallest objects (e.g., the surfaces of weapons involved) up to entire vehicles (www.gom.com).
FIGURE B1.1.7 With the help of an overhead transparency, an oriented and numbered point grid is projected onto the skin to capture the surface data by photogrammetric analysis.
© 2009 by Taylor & Francis Group, LLC
56
Since 2003, the Institute of Forensic Medicine in Bern has been using the GOM ATOS II -3D digitizer to acquire the surface of bodies, tools, and vehicles [7,8] and since 2006 has also had the GOM ATOS III -3D digitizer. External injuries of accident or crime victims, assumed weapons that were used in the deed, inflicting tools, or damages to accident vehicles are digitized and electronically stored for further investigations and analyses. Thus, it is possible to compare patterned injuries with assumed injury-causing tools and to reconstruct courses of events.
B1.2 PHOTOGRAMMETRICAL 3D DOCUMENTATION The photographic projection in a camera can be described mathematically as a central projection of the object onto the film plane. Every single object point is mapped in an image point, which is the intersection point of the projection beam and the image plane. The spatial location of a point can be determined at the intersection of the projection beams of this point from several images. In this way with the reversion of the mapping, the size and shape of the object can be reconstructed. For the three-dimensional documentation of objects, digital close-range photogrammetry is applied. With multiimage photogrammetry, the 3D coordinates of object points are computed. Therefore, the images of the object have to be taken from several angles so that every object point is visible in at least three images (Figure B1.2.1). To mark discrete points of the measuring object, three different types of reference markers are applied (Figure B1.2.2). Coded markers have a circular black-and-white pattern around the white center point. This pattern encodes the point
The Virtopsy Approach
number from 1 to 428. Thus, the automatic identification of all coded markers and the automatic orientation and calculation of the images in the 3D space are enabled. Additionally, applied noncoded circular markers are enclosed in the calculation. When the calculation is successfully completed, the 3D coordinates of all coded and noncoded markers are recorded, and the coded markers can be removed. The noncoded markers support the surface scanning and can then be used to combine surface scan data with photogrammetric images. For the fusion of body surface scan data and radiological data (multislice computed tomography and magnetic resonance imaging data), the so-called multimodality radiographic markers are applied to the body. These markers are captured by the surface scanning, and they are visible in the radiological images. Thus, they serve as combining points. A digital single-lens reflex (SLR) camera, the Nikon D2X, with a fixed focal length of 24 mm is used for photos. A ring light flash ensures constant exposure without shade effects. The camera has a resolution of 12 million pixels. The camera and the lens are not calibrated. A simultaneous calibration is processed automatically every time a series of pictures is transferred to the computer and computed with the software TRITOP. The signalization of the reference targets for the computation of the image assembly is done by using coded markers. These markers are provided in different sizes depending on the object size and the applied measuring volume. The markers are recognized in the images by the software; therefore, the computation of the coordinates runs automatically. Additionally, applied noncoded markers (reference targets) that represent discrete points of the object are defined in the same computation. These predefined reference targets then serve to orient the scan measurements during the
FIGURE B1.2.1 The measurement setup for the digital multi-image photogrammetry.
© 2009 by Taylor & Francis Group, LLC
External Body Documentation
57
FIGURE B1.2.2 The equipment of the photogrammetric documentation.
scanning process so that a full 3D model is generated step by step. A big advantage of the system is minimization of the object information. Only one reference distance has to be determined to give the right scale to the data. Therefore, a scale bar with the precisely measured distance between two coded markers can be applied to the object area. For redundancy a second scale bar is added that allows for an accuracy check.
A local coordinate system can be defined by 3D coordinates of three points. Before taking the images, the focal length of the lens has to be adjusted and should not be varied while photographing a series. For the analysis, in TRITOP every image series has to consist of at least six pictures. The images should be taken from several positions with convergent views (Figure B1.2.3). To ensure a controlled measurement of the object, every
FIGURE B1.2.3 Images series for the photogrammetric documentation of a body. The body is prepared with coded markers (the crosses), noncoded adhesive circular markers on the body, and two coded scale bars.
© 2009 by Taylor & Francis Group, LLC
58
The Virtopsy Approach
FIGURE B1.2.4 Digital photogrammetry: The result of the calculation of the image series. On the left side are the taken images, on the bottom right are the measured 3D coordinates of the coded and noncoded reference points, and on the top right the point cloud in a 3D view. The green points are the reference markers, the yellow lines are the scale bars, and the gray boxes are the camera positions.
marked object point should be visible in at least three images. To acquire the object completely, the images are taken from different elevations and with overlapping areas. With the system used in Bern, the images are transferred to the computer via a wireless LAN, and the computation is processed automatically while photographing. The results are 3D coordinates of the coded and noncoded reference points as well as the camera positions (Figure B1.2.4). These data are saved in a file. In the next step of surface scanning this reference file can be imported by the scanning software ATOS.
B1.3 OPTICAL 3D SURFACE SCANNING The GOM ATOS 3D digitizer is based on the principle of triangulation. A projection unit in the middle of the sensor head projects a fringe pattern onto the object with white light projection. Two charge-coupled device (CCD) cameras capture the fringe pattern that moves over the object’s surface in different phases (Figure B1.3.1). The sensor is connected to a high-end PC. Within seconds the software calculates the high-precision 3D coordinates of up to 4 million object points per measurement. Complex objects are digitized by taking several single measurements from different views and elevations. All these measurements are merged into a single data set. The system fully automatically determines the current sensor position by means of the reference targets and transforms the individual
© 2009 by Taylor & Francis Group, LLC
FIGURE B1.3.1 3D documentation of the external injuries of deceased with the high-resolution surface scanner, performed on the CT table. ATOS III sensor head positioned on a stand. Fringe patterns are projected onto the body surface with a white light projection; the two cameras record the fringes. Within seconds, the software calculates the high-precision 3D coordinates of up to 4 million object points per measurement.
External Body Documentation
59
is ensured. Using a calibration tool, different measurements are taken from different positions. This process takes about 5 minutes. The SO systems are mainly used for complex small parts with high demands on accuracy and data quality. With point distances of less than 0.02 mm, even the smallest details can be measured. With the ATOS III sensor in Bern, there is the possibility of configuring three different measuring areas: 1500 r 1500 mm, 500 r 500 mm, and 150 r 150 mm. There are also more configurations available. The user has to choose which are appropriate. The new ATOS III technology offers a point resolution that is four times higher in each measurement than the ATOS II system.
B1.4 PHOTOGRAMMETRY AND 3D SURFACE SCANNING OF THE DECEASED
FIGURE B1.3.2 Generation of the 3D surface model with the optical scanner: Each additional measurement is transformed fully automatically into a common object coordinate system.
measurements into a common object coordinate system. If the TRITOP is implemented first, the scanning software uses these predefined coordinates of the reference points. In this way the 3D model of the object is built up step by step. The operator can observe the digitization process continuously on the screen (Figure B1.3.2). At the end of the digitizing process, the complete object is represented by precise highresolution surface data. The measuring system can be individually adapted to different measuring volumes. The choice of the volume depends on the size of the object to be scanned and on the requirements of accuracy and resolution. With the ATOS II, three different measuring volumes in standard setup and two different measuring volumes in small objects (SO) setup were used. For the configuration of the different measuring volumes, the accordant lenses have to be mounted, and the angles of the cameras and the measuring distance have to be adjusted. Before starting the measuring process the measuring system has to be calibrated with the help of calibration tools so that the dimensional consistency
© 2009 by Taylor & Francis Group, LLC
The documentation of external injuries of the deceased is performed in the computed tomography (CT) room of the Centre of Forensic Imaging and Virtopsy. The bodies are placed on the CT table. At first, the reference targets are applied to the corpse. The noncoded markers are uniformly distributed. For the fusion with the data of the internal body, radiographic markers are fixed on the body. These markers are detected in CT and magnetic resonance imaging (MRI), and they are also captured by the surface scanner. For the photogrammetry, crosses with coded markers and coded scale bars are placed on and around the corpse. After successful photogrammetry the coordinates of the reference targets are computed and can be transferred to the scanning software ATOS. When all coded crosses and scale bars are removed, the scanning can be started. After calibration of the sensor configuration, the measurements can be taken. When the scanning of one body side is completed, a CT examination is performed. The position of the body is not changed between these procedures so that later the fusion of both data sets is easier to perform. After the surface and CT scan of one side, the body has to be turned so the whole procedure can be repeated with the back of the body.
B1.5 DIGITIZING OF INJURIES OF LIVING PERSONS When patterned injuries of a living person are documented, the scanning is processed without performing photogrammetry. The uncoded reference markers are applied to the skin where the patterned injury is located. The areas of interest are captured without predefinition of the reference markers. This is done with the first scan. During the scanning process the patient has to avoid any motion. On account of this the patient is laid down on an examination couch at the Institute of Forensic Medicine. For the documentation of the injuries one to four single
60
The Virtopsy Approach
measurements are performed, which take between 1 and 5 minutes to complete. After the scanning, a photograph of the injury is taken. With the help of the reference targets, this picture can be used to project the original color information onto the 3D model of the injury.
REFERENCES 1. Walz F.H., Wehren A., Niggi E., et. al., 1983, Automatic over-all and detail photography of autopsies. Arch Kriminol 171 (5–6), 168–72. 2. Buck U., 2001, Anwendung der Photogrammetrie bei der Polizei des Landes Baden-Württemberg. Wittwer, Stuttgart, 160 S., Deutscher Verein für Vermessungswesen Heft 2, ISSN 0940-2942, 48. Jahrgang 3. Brüschweiler W., Braun M., Fuchser H.J., and Dirnhofer R., 1997, Photogrammetrische Auswertung von Haut- und Weichteilwunden sowie Knochenverletzungen zur Bestimmung des Tatwerkzeuges—grundlegende Aspekte. Rechtsmedizin 7, 76–83.
© 2009 by Taylor & Francis Group, LLC
4. Brueschweiler W., Braun M., Dirnhofer R., and Thali M.J., 2003, Analysis of patterned injuries and injury-causing instruments with forensic 3D/CAD supported photogrammetry (FPHG): an instrument manual for the documentation process. Forensic Sci Int 132 (2), 130–38. 5. Subke J., Wehner H.-D., Wehner F., and Wolf H., 1998, Wundtopographie mittels Streifenlichttopometrie. Z. Rechtsmedizin 8 (Suppl. I), 26. 6. Subke J., Wehner H.D., Wehner F., and Szczepaniak S., 2000, Streifenlichttopometrie (SLT) A new method for the three-dimensional photoralistic forensic documentation in colour. Forensic Sci Int 113, 289–95. 7. Thali M.J., Braun M., and Dirnhofer R., 2003, Optical 3D surface digitizing in forensic medicine: 3D documentation of skin and bone injuries. Forensic Sci Int 137 (2–3), 203–08. 8. Thali M.J., Braun M., Buck U., Aghayev E., Jackowski C., Vock P., et al., 2005, VIRTOPSY—scientific documentation, reconstruction and animation in forensic: individual and real 3D data based geo-metric approach including optical body/object surface and radiological CT/MRI scanning. J Forensic Sci 50 (2), 428–424.
B2
Internal Body Documentation
CONTENTS B2.1 Conventional Radiography .............................................................................................................................................. 62 B2.1.1 Physics ................................................................................................................................................................ 62 B2.1.2 Advantages ......................................................................................................................................................... 63 B2.1.3 Disadvantages ..................................................................................................................................................... 63 B2.1.4 Practical Application in Clinical and Forensic Medicine .................................................................................. 63 B2.2 Ultrasonography (US) ...................................................................................................................................................... 63 B2.2.1 Physics ................................................................................................................................................................ 63 B2.2.2 Advantages ......................................................................................................................................................... 64 B2.2.3 Disadvantages ..................................................................................................................................................... 64 B2.2.4 Practical Application in Clinical and Forensic Medicine .................................................................................. 64 B2.3 Computed Tomography ................................................................................................................................................... 64 B2.3.1 X-Ray Computed Tomography ........................................................................................................................... 64 B2.3.1.1 Introduction ....................................................................................................................................... 64 B2.3.1.2 CT Principles and Technology .......................................................................................................... 65 B2.3.1.3 CT Applications ................................................................................................................................. 67 B2.3.1.4 Outlook .............................................................................................................................................. 69 B2.3.1.5 Acknowledgments.............................................................................................................................. 69 B2.3.1.6 References .......................................................................................................................................... 69 B2.3.2 Micro-CT............................................................................................................................................................ 70 B2.3.2.1 What Is Micro-CT? ........................................................................................................................... 70 B2.3.2.2 Spatial Resolution, Image Noise, and Radiation Dose ...................................................................... 75 B2.3.2.3 Scanner Design, Parameters, and Performance................................................................................. 76 B2.3.2.4 Micro-CT Cookbook—How to Properly Set Up a Good Scan ......................................................... 77 B2.3.2.5 Summary ........................................................................................................................................... 79 B2.3.2.6 References .......................................................................................................................................... 80 B2.4 Magnetic Resonance Imaging ......................................................................................................................................... 81 B2.4.1 Basics of MRI and MR-Spectroscopy ................................................................................................................ 81 B2.4.1.1 Short History of NMR and MRI ....................................................................................................... 81 B2.4.1.2 The Basics of the NMR Effect .......................................................................................................... 82 B2.4.1.3 Gradients Used for Spatial Encoding in Imaging and Volume Selected Spectroscopy: Image Formation........................................................................................................................................... 84 B2.4.1.4 Chemical Information ....................................................................................................................... 84 B2.4.1.5 Relaxation Times and Other Contrast Mechanisms .......................................................................... 86 B2.4.1.6 Conclusion and Outlook .................................................................................................................... 86 B2.4.1.7 References .......................................................................................................................................... 87 B2.4.2 Virtual Histology by Magnetic Resonance Microscopy .................................................................................... 88 B2.4.2.1 Introduction ....................................................................................................................................... 88 B2.4.2.2 Application of MRM to Wound Documentation ............................................................................... 88 B2.4.2.3 Mapping of Retinal Hemorrhage in Abusive Head Trauma Cases by MRM.................................... 90 B2.4.2.4 Future Prospects ............................................................................................................................... 91 B2.4.2.5 References .......................................................................................................................................... 91 B2.4.3 Nuclear Magnetic Resonance Spectroscopy in Forensic Medicine ................................................................... 93 B2.4.3.1 Introduction: Nomenclature: In Vitro versus In Vivo Applications of NMR Spectroscopy ............. 93 B2.4.3.2 In Vitro NMR Spectroscopy: High-Resolution NMR and MAS ...................................................... 93 B2.4.3.3 In Situ MRS ....................................................................................................................................... 94 B2.4.3.4 NMR Spectroscopy and Pertinent Issues in Forensic Medicine ....................................................... 94 B2.4.3.5 Potential Applications of NMR Spectroscopy to Living Persons ..................................................... 95 B2.4.3.6 The Application of MRS to the Problem of Age Determination....................................................... 98 61 © 2009 by Taylor & Francis Group, LLC
62
The Virtopsy Approach
B2.4.3.7 B2.4.3.8 B2.4.3.9 B2.4.3.10 B2.4.3.11 B2.4.3.12 B2.4.3.13 B2.4.3.14
Abuse of Drugs or Alcohol .............................................................................................................. 100 Potential Applications of NMR Spectroscopy to Diseased Persons ............................................... 104 Identification .................................................................................................................................... 104 Application of MR Spectroscopy and High Resolution NMR to the Estimation of Postmortem Intervals (PMI)........................................................................................................ 104 Cause of Death................................................................................................................................. 106 Toxicology ....................................................................................................................................... 107 Conclusion ........................................................................................................................................110 References ........................................................................................................................................110
B2.1 CONVENTIONAL RADIOGRAPHY Peter Vock Historically done at autopsy, internal documentation of the body by x-rays started soon after their invention by Wilhelm Roentgen in 1896 and the early introduction of this new type of radiation to medicine. Nowadays, a number of different imaging methods are used for internal body documentation. They are all based on one of three basic principles: reflection, transmission, or emission (Figure B2.1.1). While reflection is used in photography and ultrasound imaging, transmission is the principle of shadow imaging, radiography, and computed tomography, and emission is used in thermography, nuclear medicine, as well as magnetic resonance imaging.
l/i = (/L)2
L
i
l
B2.1.1 PHYSICS Radiography is based on x-rays, one form of electromagnetic waves of energies ranging between around 20 and 140 keV (wave lengths of 1 to 0.08 Angstroem). These rays are produced in an x-ray tube, using a tension of 20 to 140 kV between the cathode and the anode. By this enormous voltage difference, electrons emitted from the heated cathode are accelerated to reach a huge speed and energy when they hit the anode. The anode, usually made of tungsten, translates the energy of the electrons into heat and x-rays. Most of the energy is deposited in local heat, and only a small percentage is emitted as x-rays of a spectral energy range up to the maximal energy corresponding to the tension applied. X-rays have some characteristics that are important for their medical application. First of all, they irradiate linearly from the point of origin (i.e., the focus of the tube), losing their intensity by
Reflection
Transmission
Emission
FIGURE B2.1.1 The three principles of internal documentation by imaging.
© 2009 by Taylor & Francis Group, LLC
FIGURE B2.1.2 Characteristics of x-rays: X-rays propagate linearly, similar to light. They lose their intensity (I,i) by the square of the distance (l,L) from the focus of the tube. Objects (e.g., at the plane defined by l) at a distance from the image plane (defined by L) will be magnified.
the square of the distance from the tube, similar to visible light (Figure B2.1.2) with a tube close to the object, this means magnification (and often distortion) of the object. Second, they interact with materials such as tissue: At the energy levels usually used, the photoelectric effect is often predominant; an x-ray photon gives its entire energy to an electron of an atom, pushing the electron out of the atom and therefore causing ionization. This is also the reason why x-rays (as other high-energy rays) belong to the group of ionizing radiations. At energies above 100 kV, the proportion of the interaction by the Compton effect becomes large and the photo effect diminishes. The Compton effect occurs when an x-ray photon again interacts with an electron and part of the photon’s energy is transferred to the electron; however, this time some energy is left for a new photon of lower energy leaving in a different direction (Compton scatter). The third common mechanism of interaction in x-ray imaging is classical scatter, where the direction of the photon is changed without loss of energy. Interaction between x-rays and body tissue is the reason for the partial absorption of an x-ray beam on its path, which creates the transmission profile (i.e., the contrast needed). The degree of absorption increases with the wavelength of the rays and with the atomic number (more important for the photoelectric effect than the Compton effect), and it is proportional to the density and to
Internal Body Documentation
63
the thickness of the tissue. The transmission profile beyond the body can now be detected to form an image based on the locally variable transmission. To render it visible, it is caught either by a photosensitive x-ray film, by a fluorescent screen, or by a modern digital detector and is then developed, put on a screen, or printed on a hard copy. In conventional radiography, a three-dimensional object is always projected on a two-dimensional plane; therefore, the image is characterized by superimposing different structures lying in the same ray direction. It is a general rule in radiography that a second projection at a right angle to the first one should be obtained to differentiate these structures.
B2.1.2 ADVANTAGES Radiography is widely available, relatively inexpensive, and fast. In milliseconds it produces a projection image of small or large areas of the body, giving excellent information about bony, calcified, and gas-containing normal anatomy and pathology. It is therefore best used for areas of high natural contrast, such as the lung and the skeletal system. Follow-up studies will increase the sensitivity for tiny changes due to the initial pathology or to treatment-induced changes.
B2.1.3 DISADVANTAGES The two most important limitations of radiography are its projectional rather than tomographic image information and the poor contrast within soft tissues. As any type of ionizing radiation, x-rays may induce genetic changes, potentially causing fetal malformations or—even decades later—cancer.
B2.1.4 PRACTICAL APPLICATION IN CLINICAL AND FORENSIC MEDICINE Historically, radiography has been the first and—over many decades—the primary clinical imaging method and imaging application for internal documentation in forensic medicine. It is still most widely used due to its great availability. Extremity trauma to the skeleton and chest trauma are some of the most frequent indications for radiography, followed by primary disease of the lung and the pleura as well as left heart failure. It is excellent for the detection and localization of metallic foreign bodies. Dental radiography and radiography of metallic foreign bodies or vertebroplasty often help in identification. On the other hand, the recently developed modern sectional imaging methods have replaced radiography in many soft-tissue and visceral applications.
B2.2 ULTRASONOGRAPHY (US) Peter Vock
B2.2.1 PHYSICS Ultrasound waves—like sound waves—are longitudinal pressure waves; however, their frequency is beyond the audible threshold. For imaging, the frequency range of 2.5 to 12 MHz is mainly used. Ultrasound waves are produced by piezoelectric crystals (Figure B2.2.1). Subjected to pressure, these
© 2009 by Taylor & Francis Group, LLC
Principles of Ultrasonography
$ " !
" $
#
# !!
FIGURE B2.2.1 Principles of ultrasonography.
asymmetric crystals produce positive and negative electric charges on opposing surfaces, and the signs of these charges are reversed when the pressure is replaced by tension. To produce ultrasound waves, the inverse mechanism is needed; that is, an alternating electric potential is applied to provoke a cyclic expansion and contraction of the crystal. This change of size can then be transmitted to the body surface and propagates in the form of pressure waves. As sound and ultrasound waves have a constant speed of 331 m/s in air, the speed of ultrasound waves in soft tissues is relatively uniform (1540– 1560 m/s), much different from bone (3360 m/s). Within tissue, the wave propagates, causing a periodic disturbance in the direction of propagation at any point of the path, with longitudinal displacement of molecules. Minor changes of the acoustic impedance at biologic boundaries are responsible for a reflection of part of the beam, whereas the rest continues in the original direction. Major differences of impedance at surfaces of bone or gas reflect the waves more or less completely, and the beam will stop at these surfaces. This is also the reason why a gel is used to fill the gap between the probe containing the crystals and the skin to avoid any air disturbing the entrance of the waves into the body. Reflected waves, based on the constant speed in tissue, will return to the piezoelectric crystals of the transducer probe after exactly the same delay time they had needed to get to the reflecting point. The crystal, this time, will be deformed and will produce charges that can be read as an electric signal, the so-called echo. In other words, the time delay of the echo defines the depth of the reflecting boundary in the body; measurements are usually made in one plane at one time. A range of crystals and electronic modulation make it possible to get a planar interrogation and to obtain the lateral localization within the plane. By fast pulse repetition, images can be obtained in a high frequency, reaching a “real-time” movie character. This type of measurement is also called B-scan, and it is helpful to observe local temporal tissue changes, such as the pulsation of vessel walls or the motion of the diaphragm during respiration, but also to scan continuously through a volume of interest by adapting the angle of the probe and therefore the orientation of the imaging plane.
64
Ultrasound offers another measurement principle, the Doppler effect. The apparent change of frequency of ultrasound waves caused by a relative motion between the source (crystal) and the reflecting structure is called the Doppler shift. When the reflecting structure moves toward the crystal, the registered frequency increases since more waves arrive per time unit; the frequency decreases when the reflecting structure moves away from the crystal. Since the frequency shift can be quantified, both the direction and the speed of movement can be measured. The combination of Doppler measurement and B-scanning is called duplex scanning; either spectral curves show the dynamic changes of flow speed in a specific sampling volume within the gray-scale image, or the flow information is superimposed in color on the gray-scale morphologic image of the B-scan (color Doppler technique).
B2.2.2 ADVANTAGES Ultrasound offers flexible tomographic imaging in any anatomic plane. Equipment is inexpensive, both at purchase and during its use. Ultrasound waves, at the intensities used for diagnostic imaging, although they have minor biologic effects, have not been shown to cause any persisting biologic damage; this is also the reason for their widespread use in obstetric medicine. Ultrasound equipment is compact and can easily be transported to any place, be it the scene of an accident or a place in the mountains far from any civilization; in other words, ultrasound goes to the patient and not the patient to the imaging machine (in contrast to computed tomography [CT] or magnetic resonance imaging [MRI]). Furthermore, the fast dynamic measurement allows for an immediate assessment, as for instance essential in the initial management of a polytrauma victim.
The Virtopsy Approach
else, this is not possible in ultrasound since the expert will only see the frozen images that the sonographer archived.
B2.2.4 PRACTICAL APPLICATION IN CLINICAL AND FORENSIC MEDICINE Ultrasound (US) technology has made enormous progress during the past 30 years, and in outpatients and inpatients it has become the primary imaging tool, probably more often used than radiography. It offers excellent, fast information about the neck; the soft tissues of the extremities and the wall of the trunk; the visceral organs including the heart, the liver, and the kidneys; and the genital organs. It is ideal to display small fluid collections and to guide their aspiration, and similarly it guides biopsy of specific lesions. Despite its flexibility and inexpensiveness, US still is underused in forensic medicine; this can partly be explained by the training needed, the difficulty in reproducing results by an expert not involved directly with the examination, and by the difficult access in case of soft-tissue emphysema and of superficial putrefaction gas. It is likely that US will find its forensic role in investigating for metallic and nonmetallic foreign bodies, in localizing fluid collections, and in localized aspiration or biopsy in situations where autopsy is not allowed by law or not tolerated by the relatives of a body. In surviving victims, the spectrum of applications is nearly as large as in clinical medicine.
B2.3 COMPUTED TOMOGRAPHY B2.3.1 X-RAY COMPUTED TOMOGRAPHY Willi A. Kalender B2.3.1.1 Introduction
B2.2.3 DISADVANTAGES Ultrasound probes need an area of contact to the skin; since this has to be flexible and to avoid bone, it is relatively small and will always give access to a small sector or a trapezoid area of the body but not allow for a complete sectional view, as obtained by CT or MRI. Bone, foreign bodies, and any type of gas, due to their different impedance, are absolute barriers and will hide all anatomic structures behind them; this forces the ultrasonographer to change the contact point or the direction of the probe in order to look for access paths to the deep structures. Reflection at superficial boundaries is responsible for the fact that only small portions of the original waves reach deep structures; despite electronic amplification of the signal, penetration is limited. Organs of interest are often too deep in obese patients, which forces the intensity to be increased and the frequency to be decreased, with the consequence that quality gets poor; any structures far away from the skin are therefore difficult to examine. These factors also explain the extreme importance of experience in performing ultrasound examinations, and the results may differ enormously between different sonographers. Furthermore, there is no complete documentation of the volume studied: While in CT it is easy for an expert to analyze a study performed elsewhere by somebody
© 2009 by Taylor & Francis Group, LLC
X-ray computed tomography (CT) was first described in 1972. Although the initial exams were limited to the head and the brain and offered only rather crude image quality (Figure B2.3.1.1a), the method was immediately accepted for neuroradiology and achieved its breakthrough for general radiology within a few years. Its inventor, the English engineer Godfrey Newbold Hounsfield, received the Nobel Prize in physiology and medicine in the year 1979 for his important contribution to the advancement of noninvasive diagnostic radiology [1]. CT went through a decade of very rapid growth and remarkable technological developments in the 1970s. The 1980s offered far fewer innovations; the decade was characterized by the general expectation that magnetic resonance imaging (MRI) would soon replace CT. However, the advent of spiral CT scanning in the early 1990s [2] resulted in a renaissance of CT. Spiral CT meant the transition from scanning of successive single slices to continuous scanning of complete volumes. CT today allows for very fast scans of all body regions at high spatial resolution; scans are typically completed within 10 to 30 seconds at a resolution of typically 0.5 mm (Figure B2.3.1.1b). In this section, we a present and explain the principles of CT imaging and to describe the underlying technologies. Some of the more recent CT applications are presented, in
Internal Body Documentation
65
A
Δx Δy S z
y
x
FIGURE B2.3.1.2 Computed tomography—perception in slices. CT provides slice images of the human body as a matrix in digital form. A coordinate system results that basically conforms to the main anatomical axes and planes. Each picture element ( pixel) represents a volume element ( voxel) with its size corresponding to the pixel size and the slice thickness. B
FIGURE B2.3.1.1 The evolution of CT over time is well documented by respective image examples. (A) Scanning of single anatomic slices, limited to the brain and to coarse matrices at the start in 1972; (B) fast scanning of organs or body sections at high resolution enabling, for example, CT angiography of the complete craniocervical region.
particular also aspects of high-resolution imaging. The final section points to general development trends in CT, including those relevant for postmortem x-ray diagnostics (i.e., for virtopsy). We hope that all this can be fully understood without a solid background in physics, mathematics, or engineering. For readers interested in respective details, dedicated CT text books [e.g., 3,4] are recommended. B2.3.1.2 CT Principles and Technology In general x-ray radiography, images show a superposition of all structures in the object along the path from the x-ray focus to the detector. Traditionally this was carried out with x-ray film, similar to the use of film in photography, which constitutes an
© 2009 by Taylor & Francis Group, LLC
analog medium. Today, digital detectors have come into use more frequently. However, this does not resolve the inherent problem. The superposition of structure details often makes diagnoses difficult or, in many cases, impossible. In particular, no differentiation of soft-tissue structures and organs is possible. Radiography or 2D projection imaging is therefore limited to the examination of high-contrast structures such as bones. CT does not rely on a single projection image of the patient, and it does not offer analog images. It relies on calculating digital matrix images that represent single slices of the human anatomy (Figure B2.3.1.2). To achieve this goal, the respective object has to be measured from multiple directions over a range of at least 180 degrees. The measurement consists of a recording of the attenuation of the x-ray intensity along each single ray from the focus to the respective detector element. Such measurements were initially limited to 180 projections with 160 measured values each [1]. From these data, images with a quality as shown in Figure B2.3.1.1A were reconstructed; it is apparent that this is a digital image with discrete picture elements. Each picture element (pixel) represents the mean x-ray attenuation properties of a volume element (i.e., voxel); the voxel dimensions are given by the pixel size and by the slice thickness (Figure B2.3.1.2). The general CT principle—measurement of x-ray attenuation from many directions and a reconstruction of a digital slice image—holds true for all forms of computed tomography, including the most modern CT scanners as well as the high-resolution scanners mentioned herein. A typical CT examination suite is shown in Figure B2.3.1.3A, and a drawing of the actual measurement setup, the so-called CT gantry, is shown in Figure B2.3.1.3B. A high-power x-ray tube rotates around the patient. The detector, an array of single small detector elements arranged over a circle segment of typically 50 degrees, is mounted opposite it. Both tube and detector are attached to a rotating assembly and thereby move in synchrony at high rotation speeds of typically 0.3 to 1.0 second per 360-degree rotation. For two decades of CT development, this measuring assembly was connected to, for example, power supplies, computers, and cooling by cables and hoses. Therefore it was limited to single successive rotations in clockwise and counterclockwise directions. In modern designs,
66
The Virtopsy Approach
A
B Frontal View
Lateral View
x-ray tube
570
Gantry opening
Field of measurement
500
Shaped filter Fixed collimator Adjustable collimator
Center of rotation
700 435
Anti scatter collimator Detector array
Adjustable collimator Fixed collimator
FIGURE B2.3.1.3 Typical CT scanner setup. (A) CT examination room. The gantry, which typically can be tilted by ± 30n, contains the x-ray components and the measuring system. (B) The rotating system typically covers a field of measurement with 50 cm diameter.
starting in the late 1980s, slip-ring technology was developed so that the tube and detector rotate continuously for many rotations. The advent of the slip-ring technology also provided the technological basis for continuous spiral CT scanning. The principle of spiral CT scanning is shown in Figure B2.3.1.4A. While the tube and the detector rotate continuously, the patient is transported through the gantry opening at a low continuous speed. This simply means that the x-ray tube focal spot, relative to the patient, travels on a spiral or, synonymously, a helical path. While the scanning principle is easy to understand and to implement, it was a particular challenge to develop image reconstruction algorithms that allowed for uncompromised image quality. The first clinical studies with spiral CT were carried out in Bern, Switzerland, in 1989 by Peter Vock and me; one of the early studies is shown in Figure B2.3.1.4B. It was the first successful effort at 3D imaging of structures that are subject to breathing motion [2,5]. It took a few years before spiral CT was generally accepted. Today it constitutes the standard and routine measurement
© 2009 by Taylor & Francis Group, LLC
approach in modern CT. The reasons for the complete transition from slice-by-slice imaging to continuous volume imaging are convincing: r Examination times are reduced significantly by omitting the start-and-stop operation of CT scanners, the time for table transport between scans, and, in many cases, the time for breathing commands for each single scan. r There is practically no more misregistration from slice to slice due to changes in inspiratory status (patient breathing). r 3D spatial resolution, particularly in the longitudinal direction, is improved due to the continuous and partly overlapping sampling in the z-direction [3,6]. An additional technological advance was essential for the establishment of CT as the most powerful imaging modality in today’s clinical radiology: the introduction of multirow
Internal Body Documentation
67
A
Path of continuously rotating x-ray tube and detector
Direction of patient transport
Start of spiral scan 0
z, mm
0
t, s
B
FIGURE B2.3.1.4 Spiral CT allows for a fast and continuous volume scan by transporting the patient through the gantry during the examination (A). This allowed for the 3D scanning even of moving body parts (B) and is the standard scan mode on modern CT scanners today.
and array detector technology. For more than two decades CT relied on one or two linear arrays of detector elements. This meant that one or two slices could be imaged simultaneously. In 1998, four-slice imaging became available as a result of the development of the first multirow detectors. Only three years later this was extended to 16- and another three years later to 64-slice imaging (Figure B2.3.1.5). Total examination times are reduced in direct proportion to the number of slices that can be imaged simultaneously; that is, going from 4- to 64-slice imaging means that a given scan with otherwise identical scan parameters can be completed in 1/16 of the scan time. It is important to note that together with increasing the number of detector rows, the size of the single detector elements was reduced. In particular, the z-extent and thereby the slice width were reduced. Today, almost all scanners offer slice widths of less than 1 mm. Consequently, CT is the first imaging modality that offers high isotropic spatial resolution in routine scanning. Isotropic resolution means that the same level of resolution is guaranteed in all directions. A compilation of performance characteristics of CT and their development over
© 2009 by Taylor & Francis Group, LLC
time is given in Table B2.3.1.1. The increase in scan speed and in spatial resolution is most impressive; the speed and resolution are the basis for offering subminute whole-body scans at submillimeter resolution in clinical routine. B2.3.1.3 CT Applications CT offers a very broad range of applications covering all parts of the human anatomy. Its particular advantage is given by the fact that it offers both high spatial and high temporal resolution. Fast scanning also was the basis for establishing CT angiography, even for whole-body angiography, since the necessary scan can be completed in the short time for which the contrast medium bolus injected intravenously provides high contrast [3,4]. Figure B2.3.1.6 shows a compilation of respective CT examination results that can be considered typical in several respects. r All cases represent volume scans obtained with high isotropic spatial resolution of about 0.5 to 0.8 mm isotropic resolution.
68
The Virtopsy Approach
Fan beam
Cone beam
z
M.S 1 × 5 mm trot 0.75 s year 1995
4 × 1 mm 0.5 s 1998
16 × 0.75 mm 0.42 s 2001
64 × 0.6 mm 512 × 0.5 mm 0.33 s <0.33 s 2004 >2005?
FIGURE B2.3.1.5 The development of multirow detectors allowed for faster and faster scanning by going from single-slice to multislice acquisition. It is unclear how much further this development will continue; 64-slice scanning presently represents the state of the art and, in fact, a mature and reliable solution.
r The scan data and the reconstructed and processed images are all digital. They are transferred to arbitrary stations in the hospital and are archived in digital format. Film or paper copies are the exception but are typically used to document the diagnostic findings. r Visualization of results is not limited to the typical transverse slice but is shown at arbitrary orientations and 2D and 3D views. (It should be noted that presently the diagnostic evaluation of such scans is mostly done interactively at a viewing station. Colored 3D displays are not the primary TABLE B2.3.1.1 Performance Characteristics of CT Scanners in a Comparison from 1974 to 2004 1974
1984
Minimum Scan Time 300 s 5–10 s Slice Thickness 13 mm 2–10 mm Spatial Resolution 3 Lp/cm 8–12 Lp/cm 57.6 kB 1 MB Data per 360n Scan Data per Spiral Scan — — Image Matrix 80 r 80 256 r 256 Power 2 kW 10 kW
1994
2004
1–2 s 1–10 mm 10–15 Lp/cm 1–2 MB 24–48 MB 512 r 512 40 kW
0.33–0.5 s 0.5–1 mm 12–25 Lp/cm 10–100 MB 200–4000 MB 512 r 512 80 kW
Note: The values represent some typical data for high-performance scanners.
© 2009 by Taylor & Francis Group, LLC
FIGURE B2.3.1.6 Typical examples for fast clinical volume scanning. 3D evaluation and rendering are the standard today.
Internal Body Documentation
69
“Routine head” “Ultrahigh res.”
10 lp/cm 12 lp/cm
20 lp/cm 18 lp/cm 16 lp/cm
14 lp/cm
FIGURE B2.3.1.7 High-resolution CT allows for isotropic resolution up to typically 0.4 mm. In-plane resolution can be increased even further (lower left), but at the expense of higher noise.
image for the diagnosis but are a very useful and often convincing way of rendering the diagnosis to the referring physicians and even to the patient.) The whole-body exam on the left of Figure B2.3.1.6 was obtained in 28 seconds. It is important to note that this is achieved without compromising isotropic spatial resolution. This feature is also an indispensable requirement for many other applications where slice images have to be chosen at arbitrary orientation or for virtual endoscopy, such as virtual colonoscopy. The high-resolution capability of modern CT, which is important for postmortem imaging, is demonstrated in Figure B2.3.1.7 for a resolution pattern that allows for a more objective assessment than clinical images. High-resolution scan and reconstruction modes deliver typically 0.5 mm or better; in the case displayed even 0.4 mm, corresponding to about 12 line pairs per mm, are reached both in the transverse plane and in the longitudinal direction. Respective protocols are typically used for bone and lung scans—that is, for examinations or structures that offer high contrast to their surroundings. The most demanding task for all imaging modalities is the imaging of the heart, in particular of the coronary vessels. This has become a strong clinical application of modern CT just recently. The respective technology and methodology became available in the late 1990s [3,7]. It has been augmented just recently by the advent of so-called dual-source CT scanners, which employ two complete measuring systems mounted on one gantry [3,8]. Such systems offer the particular advantage that the minimum scan time is simply reduced by a factor of two. Since images can be reconstructed from partial scans covering only 180 degrees of data, the minimum scan time in this case is less than 100 ms. CT has always excelled with respect to the display and diagnosis of anatomy and its morphology. Recently new approaches to functional imaging have become available. The measurement of brain perfusion by dynamic CT is one important clinical example. Most important for oncological applications was the development of systems, which combine a CT and a positron emission tomography (PET) system back
© 2009 by Taylor & Francis Group, LLC
to back in a single apparatus [9,10]. These so-called PET/CT combination scanners allow for standard CT examinations with the PET results representing lesion metabolism or other functional parameters superimposed onto the CT images. B2.3.1.4 Outlook Although CT is based on mature technology and is solidly established in clinical radiology, there are quite a number of developments to be expected in the future. These will not be driven predominantly by advances in technology but mostly by clinical demands and workflow considerations. Accordingly, they will not necessarily be of importance for virtual autopsy. One trend, however, may prove to be of high practical value for forensic medicine: the development of higher-resolution CT imaging capabilities using C-arm systems equipped with so-called flat detectors (Figure B2.3.1.8A). The first respective efforts date back to about one decade ago; the technology is presently entering clinical routine with a focus on intraoperative imaging and on supporting interventional procedures [11]. Flat detectors, originally developed for radiography, do not rely on film but allow for direct digital readout using a scintillator input screen that is coupled directly to a photodiode array. They exhibit detection efficiency and general performance characteristics slightly lower than dedicated CT detectors. However, they offer significant advantages, in particular higher spatial resolution due to the smaller size of the detector elements. This is demonstrated by the examples in Figures B2.3.1.8B and B2.3.1.8C: 0.2 mm structures, corresponding to 2.5 line pairs per mm, are clearly resolved, the trabecular network of a femur specimen just the same. Flat-detector technology also presents the basis for another development in the field of CT: dedicated scanners for high-resolution imaging of small objects and tissue samples. Since resolution levels of typically 10 to 100 micrometers are achieved as demonstrated by Figure B2.3.1.9, this novel technique is generally called micro-CT. It is dealt with in detail in a separate chapter in this volume. Flat-detector CT in general and micro-CT in particular will make CT of even higher value for virtual autopsy. B2.3.1.5 Acknowledgments The majority of the figures were taken from Computed Tomography [3] with friendly permission from the publisher. B2.3.1.6 References 1. Hounsfield G.N. 1973. Computerized transverse axial scanning (tomography). Part I. Description of system. Br. J. Radiol. 46, 1016–1022. 2. Kalender W.A., Seissler W., Klotz E., and Vock P. 1990. Spiral volumetric CT with single-breathhold technique, continuous transport, and continuous scanner rotation. Radiology 176, 181–183. 3. Kalender W. 2005. Computed Tomography, 2d ed. Erlangen: Publicis. 4. Prokop M., Galanski M., Schaefer-Prokop C., and van der Molen A.J. 2007. Ganzkörper-Computertomographie. Spiral- und Multislice-CT, 2d ed. Stuttgart: Georg Thieme.
70
The Virtopsy Approach
A
B
C
FIGURE B2.3.1.8 C-arm-based x-ray systems provide improved performance due to the introduction of electronic readout flat detectors (A) including the option for CT scanning. Spatial resolution is increased further, here to 0.2 mm, as shown both on bar pattern phantom (B) and on a femur specimen (C).
5. Vock P., Soucek M., Daepp M., and Kalender W.A. 1990. Lung: Spiral volumetric CT with single-breathhold technique. Radiology 176, 864–867. 6. Kalender W.A. 1995. Thin-section three-dimensional spiral CT: Is isotropic imaging possible? Radiology 197, 578–580. 7. Kachelrieß M., Kalender W.A. 1998. ECG-correlated image reconstruction from sub-second spiral CT scans of the heart. Med. Phys. 25, 2417–2431. 8. Flohr T.G., McCollough C.H., Bruder H., Petersilka M., Gruber K., Süß C., et al. 2006. First performance evaluation of a dual-source CT (DSCT) system. Eur. Radiol. 16, 256–268. 9. Townsend D.W., Beyer T., and Blodgett T.M. 2003. PET/ CT scanners: A hardware approach to image fusion. Sem. Nucl. Med. 33 (3), 193–204.
© 2009 by Taylor & Francis Group, LLC
10. Von Schulthess G. 2003. Clinical Molecular Anatomic Imaging: PET, PET/CT, and SPECT/CT. Philadelphia: Lippincott, Williams & Wilkins. 11. Kalender W.A. and Kyriakou Y. 2007. Flat-detector computed tomography (FD-CT). Eur. Radiol. Epub ahead of print (PMID: 17587058).
B2.3.2 MICRO-CT Marek Karolczak and Willi Kalender B2.3.2.1 What Is Micro-CT? Micro-CT can be seen as a natural extension to medical x-ray computed tomography (CT). What makes it distinguishable from clinical (human/“macro”) CT is its spatial resolution,
Internal Body Documentation
71
Lung
Liver
Heart FIGURE B2.3.1.9 Micro-CT allows for even higher spatial resolution as demonstrated by the whole-body angiography of a mouse at better than 100 Mm and by the in vitro scans of excised organs at better than 10 Mm. (Vascular contrast agent provided as a courtesy by the Institute of Forensic Medicine, Bern, Switzerland.)
which ranges (depending on the application) from submicrometers to a few hundred micrometers. So far no clear definition of micro-CT exists; however, an intuitive and justified one would be “CT with spatial resolution better than 100 Mm.” A significant difference between micro-CT and clinical CT is that samples are smaller, adjusted to resolution. (Table B2.3.2.1 and Figure B2.3.2.1) CT was created in the early 1970s and soon became the imaging modality of choice in a clinical environment. In a relatively short time a number of manufacturers—such as Elscint, GE, Philips, Picker, Siemens, and Toshiba (in alphabetical order), to name a few—established their presence on the market. Micro-CT came out some 10 years later [7,32] as people discovered the potential of CT for investigating small objects, (e.g., biological samples) [2]. The technology was there, and components like x-ray tubes with a very small focus (microfocus) were commercially available [3]. Several research groups around the world began working on the development of micro-CT scanners. Symptomatically these early works and the most significant and spectacular achievements did not take place at the labs of the “giants” but were done by small groups.
© 2009 by Taylor & Francis Group, LLC
Micro-CT has been applied in various fields of medical and biological research such as drug development and implants. In the industry, micro-CT has become a significant part of the nondestructive testing (NDT) methods, where increasing miniaturization of electronic and mechanic components demands the development of methodology enabling the examination of inner structures of these components in microscale. In such applications microtomography is the method of choice. B2.3.2.1.1 Micro-CT versus Clinical (Human) CT From the very beginning micro-CT scanners differed significantly from their “big brothers”—clinical CT—not only in their dimensions and resolution but also in the scanner design and concept of the scan geometry. What in clinical CT has evolved over the years from a single slice to multirow technology and is in the near future expected to expand to so-called true cone-beam geometry has been present in micro-CT from its early birth. Thus, volumetric scanning was possible with a single gantry rotation around the sample.
72
One significant distinguishable difference is that in micro-CT scanners the x-ray source and detector assembly are often statically mounted and the sample rotates. This “reversed” setup is easier to build, is mechanically more stable, and offers the possibility to vary the geometry, hence providing adjustable spatial resolution. On the other hand, mechanical accuracy and stability (e.g., thermal) of the scanner, which is required to achieve micrometer resolution, push design complexity to the extent not known in clinical scanners. X-ray sources used in micro-CT scanners are orders of magnitude weaker than those in clinical scanners, and this leads to prolonged scan times and noisier images. However, miniature focus and requirements for its stability justify high costs of these components. Probably in all micro-CT scanners only commercially available radiation detectors (i.e., not custom designed) are used to acquire projections. These detectors are not optimized for CT imaging and therefore impose limitations on scanner performance (e.g., scanning speed). Detectors in clinical CT scanners are, on the contrary, highly optimized for the purpose they serve. Software for data collection and reconstruction (creation of volumetric data) is at least equally complex as in clinical scanners or, given the size of micro-CT volumes (10243 as opposed to typically 5123 in clinical scanners), even more demanding. Nevertheless, specialized hardware accelerators for speeding up processing are seldom implemented in micro-CT scanners; therefore, data processing usually takes longer than in clinical scanners. All that explains why high-end, high-performance micro-CT scanners are not significantly cheaper than an average clinical scanner. With respect to image quality, one has to accept that in micro-CT the price to pay for increased resolution is noisier images and higher radiation dose (this topic will be discussed later). On the other hand, thanks to area detectors with (usually) square pixels, micro-CT scanners produce images with isotropic spatial resolution. In clinical scanners the in-plane resolution is typically better than resolution along the rotation axis (the z-axis). Traditionally in clinical CT one speaks about “slice thickness” and spatial resolution, which may differ from each other in the order of magnitude. On the contrary, in volumetric scanning using area detectors, as is the case in micro-CT, there is no reason why these two quantities should be different. Detector elements are square, so one should think of the volume as an isotropic, homogeneous entity, represented as an set of cubic voxels. Slice thickness should be seen as the depth of a voxel and equal to the inplane pixel size. B2.3.2.1.2 Micro-CT Scanners Are Not Equal The spectrum of micro-CT applications is probably wider that that of clinical CT, although the following selection is far from being complete. Due to variety of applications and variety of scanner configurations and designs, micro-CT can be categorized into several classes. Depending on the criteria the following groups can be defined.
© 2009 by Taylor & Francis Group, LLC
The Virtopsy Approach
B2.3.2.1.2.1 Application r Medical or biological in vivo, where scan time and radiation dose play a vital role: In particular, small animal imaging (SAI) has recently gained in importance (e.g., in the pharmaceutical industry) and in basic research, where mice, rats, and other laboratory animals are being used to model humans [4,16,24,30]. (Figure B2.3.2.2.) r Medical and biological ex vivo and in vitro, where scan time is usually not critical: It may play a significant role only if the specimen is unstable and may require rapid examination (e.g., soft tissue); radiation dose is usually of secondary importance (assuming no destructive influence of radiation on the specimen) [5,17,18,22,28,34–37]. (Figure B2.3.2.3 and B2.3.2.4.) r Nonmedical applications (e.g., industrial samples undergoing NDT), where scan time usually does not play any role (as long as the specimen remains stable) and the radiation dose is unimportant (again, assuming no destructive influence of radiation on the specimen) [6,21,25]. (Figure B2.3.2.5.) B2.3.2.1.2.2 Gantry Design r As in clinical CTs the gantry can rotate around the object, which remains at rest on a (usually motorized) patient table; this mode of acquisition is particularly suitable for in vivo examinations of sedated small animals; the disadvantage of this arrangement is that the scanner geometry (i.e., distances between the source, or tube) and the detector are usually fixed, disabling any magnification and spatial resolution variability [7]. r In a reversed arrangement, which is used in industrial applications and in ex vivo and in vitro medical examinations, the object is placed on a rotating table between statically mounted x-ray tube and detector; usually the scan geometry (i.e., the distances between the tube, specimen, and the detector) are variable, providing adjustable spatial resolution [8,31]. B2.3.2.1.2.3 Radiation Source r Synchrotron, where radiation comes from a particle accelerator (synchrotron) as known from nuclear physics: This form of radiation is monochromatic—that is, all radiation photons carry (almost) the same energy, what has a positive impact on the image quality (increased contrast). The beam intensity is significantly higher than in the x-ray tubebased scanners, resulting in shorter scan times; however, the disadvantage of this technique is its cost and lack of portability, as such scanners must be located close to an accelerator—this makes this technique of marginal importance [9].
Internal Body Documentation
73
15 cm
5 cm
5 mm
Standard CT ~ 1 mm
High resolution CT 100 μm–1 mm
Micro CT < 100 μm
FIGURE B2.3.2.1 Example of the very same spine sample scanned at different resolution.
r X-ray tube, as it is in clinical CT scanners: These systems are more flexible, but radiation is not as “clean” as the synchrotron radiation—it is polychromatic (i.e., photons carry different energy) with all the consequences on the image quality.
FIGURE B2.3.2.2 Mouse in vivo scan: (left) coronal view; (right) maximum intensity projection (MIP).
© 2009 by Taylor & Francis Group, LLC
Another weakness compared with synchrotron radiation is lower photon flux due to limited thermal capacity of the tube, resulting in prolonged scan time (the tube power in watts is of the order of the focus size in micrometers) [3].
FIGURE B2.3.2.3 Examples of in vitro scans of medical samples: (top) bone sample with a knife blade outline fitted to the shape of the wound; (bottom) rat knee (with a 3D surface view).
74
The Virtopsy Approach
FIGURE B2.3.2.4 Examples of in vitro scans of biological specimen: (left) ex vivo mouse kidney scan using contrast agent (15 Mm resolution); (right) deep sea coral (15 Mm resolution).
In practice, only x-ray tubes have been established as a standard source for micro-CT scanners (similar to clinical CT), whereas synchrotron-based scanners remain a marginal niche, mainly for research purposes, and are bound to accelerator labs. B2.3.2.1.2.4 Beam Geometry Parallel acquisition geometry means simpler (and therefore faster) data processing and artifact-free images if circular-
FIGURE B2.3.2.5 NDT samples: (top) 3D rendered view of a stent; (bottom) sagittal view of an electrolytic capacitor.
© 2009 by Taylor & Francis Group, LLC
scan trajectories are used; however, it requires detectors of high spatial resolution to achieve high spatial resolution of the scanner as a whole; parallel geometry is typical for synchrotron-based scanners. Diverging-beam geometry (fan and cone) enables “magnification” of the specimen, which is projected onto the detector surface; detectors with larger pixels (lower spatial resolution) can be used for scanning as they do not limit resolution of the scanner as a whole anymore; variable magnification allows for scanning objects of various sizes and with various pixel and voxel sizes. In practice, only diverging-beam geometry is being used in commercial micro-CT scanners, leaving the parallel case to marginal, experimental devices. As depicted at the bottom of Figure B2.3.2.6, the x-ray source (its focus marked with s) projects object voxel w onto the detector surface. The source s must be small enough compared with w to achieve a sharp image on the detector; otherwise (s significantly larger than w), blurred image w` would be created. However, the extent of blurring is high only if the sample is located close to the focus (high magnification). If the sample is placed close to the detector (low magnification), then blurring would be negligible; hence, the two scanner setups are referred to as follows: 1. Magnifying geometry (MG): when the object is located close to the tube; the tube focus determines the best achievable resolution; tubes with small focus have limited power so that scan times are prolonged (minutes to hours). 2. Contact geometry (CG): when the object is located close to the detector; the tube focus can be large, as the scanner spatial resolution is then determined by
Internal Body Documentation
75
Parallel Beam D
w
w
Detector
Diverging Beam (fan, cone) w Source w´ s
a
b
Detector
FIGURE B2.3.2.6 Illustration of the parallel-beam (top) and fanand cone-beam (bottom) geometry.
the detector resolution; here the tube may irradiate higher power so that scans can be accomplished in shorter time; this setup is usually used in high speed, in vivo scanners with moderate resolution. MG scanners offer orders of magnitude better spatial resolution (1–10 Mm) than CG, since it is technologically easier to manufacture microfocus tubes than detectors with micrometer resolution. On the contrary, CG scanners are well suited for high-speed scanning (e.g., with in vivo examinations), due to the allowance of higher tube power. Here moderate spatial resolution of typically 50–200 Mm can be achieved. B2.3.2.1.2.5 Scan Mode and Trajectory As in clinical CT, in micro-CT the following two scan trajectories are potentially available: r Circular r Spiral/helical Since in most micro-CT scanners area detectors are used to acquire projections, the entire volume of interest may be covered in a single gantry rotation around a patient. For objects in which the length does not exceed the length of the volume of measurement (VOM), circular-scan trajectories are used. These trajectories are simpler and require shorter scan times
© 2009 by Taylor & Francis Group, LLC
and lower radiation doses. It is irrelevant whether the gantry rotates around the specimen or the specimen is being rotated (reversed CT arrangement). Commonly in both gantry concepts simultaneous gantry and patient table movement can be implemented, allowing for spiral/helical acquisition. Despite large coverage of the specimen volume in a single gantry rotation, spiral mode may be preferred in some cases, as it provides artifact-free images. In circular trajectories image artifacts are generated resulting from the fact that cone-beam projections collected in a single circular scan provide, mathematically speaking, insufficient data for error-free volume reconstruction. Hence, only approximate reconstruction is possible. The extent of the artifacts depends on the cone angle—larger produce stronger artifacts. Cone-beam spiral CT reduces or eliminates these artifacts. Although the problem of insufficient data in circular-scan trajectories concerns both micro-CT and human CT, it does not play any significant role in clinical scanners. Due to the limited extent of the VOM along the rotation axis (the z-axis) in clinical CT spiral scanning is unavoidable, so projections acquired there meet the sufficiency condition and lead to artifact-free volumes. On the other hand, micro-CT scanners are commonly equipped with area detectors, so circular trajectories are the natural choice for them; therefore, images suffer from (usually subtle) artifacts [2,26]. (Figure B2.3.2.7.) Over the years, several reconstruction algorithms, both approximate and exact, have been developed. Among them, the most common is the approximate algorithm for circular scans proposed originally by Feldkamp. B2.3.2.2 Spatial Resolution, Image Noise, and Radiation Dose Almost all micro-CT scanners—both commercial and experimental—with rotating specimen (reversed gantry design) and a few of those with rotating gantry offer variable magnification, meaning the field of measurement (FOM) is adaptable to the specimen size. The advantage of this feature is the possibility of reaching the maximum achievable spatial resolution for a given sample size. However, increased spatial resolution means that the voxel size in the specimen becomes smaller and, as a consequence, the number of x-ray photons passing through the voxel lowers. This in turn has a negative impact on the image noise, which is (in a well-designed scanner) predominated by photon statistics. The lower is the number of photons penetrating the voxel, the higher is the statistical noise and the noisier appear (subjectively) the images. To compensate for this effect, the photon “density” per voxel has to be increased, which in turn increases radiation dose deposited in the object. This side effect is of importance, or sometimes unacceptable, when too high a dose might damage the sample (particularly the case for living objects). It is therefore vital to keep the dose below the allowed, tolerable level and to accept inevitable limitations in the image quality incurred. Photon statistics—that is, the number of photons impinging upon the specimen per scan—can be increased in several ways:
76
The Virtopsy Approach
r X-ray source power can be increased. This is possible only in tubes with large focus, while microfocus tubes have limited power. The focal spot in the anode material of the tube can absorb only limited power; otherwise, the anode material would melt. Consequently, this approach is feasible only in scanners with moderate spatial resolution, based on the CR principle. r If the tube power cannot be increased (due to the limitation imposed by the focus size), the exposure time per scan has to be extended. In some cases this might be undesired if rapid, time-unstable specimens are to be scanned (e.g., objects scanned in vivo). r Alternatively, the tube–detector distance can be reduced (short scanner geometry). Shorter distance to the source means increased cone angle at which the specimen is “seen” from the tube and therefore increased photon flux through voxels. The beam intensity per voxel follows the square distance law—that is, an n-fold reduction in the distance causes n2 increase in the photon flux. This approach has one drawback: Increased cone angle causes stronger image artifacts compared with long geometry (i.e., long tube–detector distance) if circular acquisition trajectory is used (Figure B2.3.2.7).
Circle cone-beam
Spiral cone-beam z
z
x
x 5° cone
11° cone
30° cone
Circle
Spiral
Although intuitively not easy to explain, it can be shown that spatial resolution and radiation dose are related to each other with the fourth power. In order to double spatial resolution for a given sample and to achieve the same level of noise per image voxel, radiation dose would have to be increased by a factor of 16 (24 16). Doubling radiation dose would allow an increase in resolution of ca. 20% only (at the same image noise level). If dose cannot be increased beyond some limit without sacrificing the specimen (e.g., living animal), then either spatial resolution has to be lowered or higher noise level must be accepted [10,12,19,20,23,33]. B2.3.2.3 Scanner Design, Parameters, and Performance In order to give the reader a better hands-on feeling of a real device, two representative, modern scanner examples are presented here: (1) a high-resolution scanner (Institute of Medical Physics, University Erlangen, Germany [2]); and (2) a high-speed, in vivo scanner (VAMP GmbH, Germany [7]). B2.3.2.3.1 High-Resolution In Vitro/NDT Scanner The high-resolution scanner (Figure B2.3.2.8, top) is built around a microfocus x-ray tube and a two-dimensional, low-noise area detector. The tube beams continuously, and a mechanical shutter placed in front of it limits exposure. The tube has a small focal spot size selectable among 5, 20, and 50 Mm, wide high-voltage range (40 to 150 kV), and reasonable power (up to 10 W at 5 Mm focus size). The detector is built around a charge-coupled device (CCD) chip with 10242 pixels of 57 Mm2 size each, cooling to –40 degrees (to reduce electronic noise) and an x-ray absorber (phosphor screen). The x-ray tube and the detector are fixed to the scanner bed, and the sample is placed between them on a rotation table. All scanner components are remotely controlled from a computer. The scanner has variable geometry (tube-to-object distance, tube-to-detector distance, object magnification)—that is, adjustable resolution adaptable to the sample size. The parameters result in the following overall performance of the scanner: r Variable field of measurement of 1–50 mm diameter and equal height (length along the z-axis) r Variable voxel size from 5 Mm to 100 Mm r Projection size selectable between 5122 and 10242 r Scan times ranging from minutes to hours r Spatial resolution better than 10 Mm B2.3.2.3.2 High-Speed In Vivo Scanner
FIGURE B2.3.2.7 (Top) Circular- and spiral/helical-scan trajectories; (bottom) artifacts in a sagittal view of a multiple-disk (“Defrise”) phantom scanned at various cone angles using circular and spiral/helical trajectories.
© 2009 by Taylor & Francis Group, LLC
The high-speed scanner (Figure B2.3.2.8, bottom) is built around a moderate-focus-size x-ray tube and a two-dimensional high-speed area detector. The tube beams continuously and has high voltage ranging from 30 to 65 kV and moderate power (up to 50 W). The detector is built around a photodiode array chip with 10242 pixels of 50 Mm2 size each, without cooling and with x-ray absorber (phosphor screen).
Internal Body Documentation
77
Microfocus x-ray tube 2D detector
The scanner is fully radiation-shielded, so it can be used in any working environment. Thanks to relatively low-tube voltage range (up to 65 kV), shielding is not very heavy, making the scanner lightweight (less than 250 kg), so it can be placed on a laboratory table. B2.3.2.3.3 Market Overview Table B2.3.2.2 lists, in alphabetical order, manufacturers offering commercial micro-CT scanners. It is definitely incomplete yet representative of the market. The market has matured from infancy but is still dynamically evolving, and new companies continue to emerge with their products. Some of the companies listed in the table specialize in medical or in vivo applications; others concentrate on industrial applications; and some cover all applications. Some manufacturers, when describing their scanners, often quote voxel size as performance measure instead of spatial resolution. These two quantities are not equal, and this information may be misleading for the potential user. The best achievable spatial resolution in a given scanner is worse than the voxel size (in the order of twice the voxel size), so scanner specifications should be interpreted with caution. The recommended way for selecting a scanner optimally suited to the task is to objectively verify its performance with test objects (e.g., phantoms) [11]. B2.3.2.4 Micro-CT Cookbook—How to Properly Set Up a Good Scan
FIGURE B2.3.2.8 Photographs of two modern micro-CT scanners: (top) high-resolution scanner [2] with rotating sample; (bottom) high-speed in vivo scanner [7] with rotating gantry.
We will now consider hypothetical cases of two samples, one of which should be scanned at high resolution (no dose limitations) and the other at high speed (where resolution is of secondary importance). We will demonstrate how to properly choose scan parameters to get the best possible image quality. For the analysis we assume a scanner with variable geometry/resolution and adjustable tube high voltage. The scanner is equipped with a detector having two projection matrix settings: 10242 and 5122 with 50 Mm2 and 100 Mm2 pixel size, respectively (ca. 50 r 50 mm2 detector active area). B2.3.2.4.1 Sample and Voxel Size
Here the scanned object lies on a motorized table, and the tube–detector assembly rotates around it. All scanner components are remotely controlled from a computer. The scanner has fixed geometry (tube-to-object and tube-to-detector distances), so no resolution adjustment is possible. The aforementioned parameters result in the following overall performance of the scanner: r Fixed volume of measurement of 40 mm diameter and approximately 40 mm length r Fixed voxel size selectable between 40 and 80 Mm (depending on the projection matrix) r Projection size selectable between 5122 and 10242 r Scan times ranging from a few seconds to minutes r Spatial resolution better than 80 Mm
© 2009 by Taylor & Francis Group, LLC
Apart from the three correlated factors mentioned in Chapter B2.3.2.2 in this volume—resolution, noise, and dose—that directly influence the image quality, another aspect of the scan setting has to be taken into account when planning a scan: the sample (specimen) size. The scanner geometry requires certain specimen dimensions. Whereas the length of the sample is of no effect, its diameter is critical. If the diameter exceeds the width of the VOM determined by the angle of the x-ray beam, artifacts will occur in reconstructed images. In clinical CT this effect is known as the “fat man problem” and can be corrected only to some limited extent. Specimen exceeding the length of the VOM can be sequentially scanned at different patient table positions and partial subvolumes can be concatenated. The sample of a diameter D mm is projected onto the detector matrix of width W pixels and pixel size of P Mm (implies the detector physical width of WxP mm). The
78
The Virtopsy Approach
TABLE B2.3.2.1 Comparison of Clinical and Micro-CT Scanners Clinical (Human) CT
Micro-CT
Gantry
Fan-beam radiation geometry, multirow detector, patient at rest (gantry rotates around the patient)
Best spatial resolution Specimen diameter X-ray tube
0.25 mm (in-plane) Typically ~1.0 mm Max. 70 cm High power, large focus (~1 mm)
High voltage Number of slices per gantry rotation Scan time
80–140 kV Typically 64 Experimentally up to 256 Typically 1–40 s
Radiation dose Image noise/low contrast detectability
5–50 mGy per examination Low, enables soft-tissue differentiation
detector height is of secondary importance, but usually micro-CT detectors are square (or close to square). The resulting magnification factor determining the tube-to-sample and sample-to-detector distances must not exceed WxP/D. The voxel size in the sample must be larger than D/W. Hence, the best spatial resolution that can be achieved for a given sample would be slightly worse than the voxel size (at least for high-contrast objects). Should the anticipated resolution be unsatisfactory for the examination, then either the sample would have to be thinned (reduced D) or the detector matrix W increased (e.g., doubled) at the same VOM size. The first option might not always be possible if the sample needs to be scanned intact; the second, on the contrary, has a drawback of a significantly increased amount of computer data to
Cone-beam radiation geometry, area detector, patient at rest (in vivo scans), or Cone-beam radiation geometry, area detector, gantry at rest, patient rotates (reversed CT) 1 Mm–100 Mm (isotropic) 1–100 mm Medium power, medium size focus (50–200 Mm) Low power, micro-focus (1–50 Mm) 20–450 kV Typically 500–1000 Up to 4000 commercially available A few seconds (medium resolution, high speed in vivo) Minutes to hours (high resolution) mGy (in vivo) to Gy (in vitro, NDT) Higher than in clinical CT Very high at ultrahigh resolution
be processed and prolonged scan time (if the level of noise should be preserved). r Case 1—in vitro high-resolution scan: A sample of 10 mm diameter needs to be scanned with at least 20 Mm spatial resolution. We set the voxel size to 10 Mm, which is twice smaller than the required resolution. This results in a projection matrix size of 10242 pixels (10 mm diameter/10 Mm voxel size 1000). Considering a 50 mm2 detector area, this would result in 5:1 magnification geometry (50 mm/10 mm). To get a sharp image of the object (at 10 Mm voxel), a low-power microfocus x-ray tube with focus size smaller than 10 Mm would be needed.
TABLE B2.3.2.2 Manufacturer
Location
Aloka DRGEM GammaMedica
Japan Korea United States
General Electric/EVS Phoenix Rapiscan Scanco Medical AG Siemens/ImTek SkyScan Stratec Toshiba VAMP
United States Germany United States Switzerland United States Belgium Germany Japan Germany
© 2009 by Taylor & Francis Group, LLC
Type In vivo In vivo, NDT In vivo, NDT, multimodality (PET/SPECT) In vivo NDT NDT In vivo, in vitro In vivo, multimodality (PET) All applications, dual source NDT NDT In vivo, dual source
Web Site http://www.aloka.co.jp http://www.drgem.co.kr http://www.gammamedica.com http://www.gehealthcare.com/euen/fun_img/products/pre-clinical/index.html http://www.phoenix-xray.com http://www.rapiscan.com http://www.scanco.ch http://www.medical.siemens.com http://www.skyscan.be http://www.stratec-med.com http://www.toshiba-itc.com/cat/en/prod02.html http://www.vamp-gmbh.de
Internal Body Documentation
r Case 2—in vivo high-resolution scan: A sample of 30 mm diameter needs to be scanned at 150 Mm spatial resolution. The voxel size needs to be smaller than 75 Mm (150 Mm/2) so that in this case smaller projection matrix of 5122 pixels can be selected (30 mm/75 Mm 400) and magnification would lie around 1,3:1 (100 Mm detector pixel/75 Mm voxel). In this case, since the object would be placed close to the detector, we could afford a tube with larger focus (50–100 Mm) and higher power, resulting in shorter scan time. B2.3.2.4.2 High Voltage In clinical CT patients are commonly scanned (depending on the examination) at 100–140 kV and in special cases (e.g., infants) at 80 kV [12]. These settings result from long-term experience acquired over decades of CT application in clinical routine. However, these values should not be directly carried over to micro-CT. Let us consider the following thought experiment: A given specimen, composed of material having certain x-ray “stopping power” (radiation attenuation coefficient), needs to be examined in a micro-CT scanner. We send N x-ray photons onto it, and some of them will be absorbed in the object. We count N` photons leaving the sample on the other side and arriving at the detector. If too many photons were absorbed in the specimen we would record only very few of them, so the statistical measurement error (noise) at the detector would be high. If, on the contrary, only a few were stopped in the sample, then almost all primary photons would reach the detector. Again, we would have very limited information about the object and high statistical error of the measurement, as too few photons would interact with the sample. We conclude that there must be some “golden middle,” some optimum for the x-ray absorption value, at which signal-to-noise ratio (SNR) reaches its maximum. Indeed, it can be shown that the optimum SNR is achieved if D * M ~ 2.22 (D denotes the object thickness and M the linear attenuation coefficient) [9]. This corresponds to approximately 10:1 x-ray absorption in the object. We can make use of this formula in setting up our hypothetical scan. Of course, the object thickness D cannot be altered (the sample has its definite shape and size), but we can influence the value of M. Attenuation coefficient M depends on the energy of x-ray photons, so through proper selection of the tube high voltage we can tune the x-ray beam energy to that level at which M in the sample is close to the SNR optimum. For example for a 20–40 mm diameter water-equivalent objects (e.g., soft tissue, small animals) the optimum kV setting would lie around 30–40 kV, and for stronger absorbing samples of the same size the setting would be higher (bones: 60–80 kV; metals: > 100 kV) [13]. B2.3.2.4.3 Scan Time The absolute scan time is not easy to predict as it depends on many factors. However, some estimation can be done considering the previously selected projection matrix size. In order
© 2009 by Taylor & Francis Group, LLC
79
to achieve isotropic resolution in the entire volume, a sufficient number of projections must be acquired. Mathematically speaking, the number of projection views must be related to the size of the acquisition matrix expressed in pixels. Switching projection matrix from 5122 to 10242 (as in our hypothetical scans) would require doubling the number of projections. In practice, the scan time would not only double but will (typically) also increase by the factor of four to eight. This happens due to the fact that in realistic detectors the readout rate (number of images per time unit they can deliver) is inversely proportional to the matrix size. If P projections per second can be acquired at a smaller matrix, then only P/2 per second (or even less) are available at larger matrix. Coming back to our example cases, if Sample 1 could be scanned in T seconds (or any arbitrary time units) using 10242 projection matrix, then scan of Sample 2 would be completed in a quarter of this time (twice smaller matrix Æ double the detector readout speed and half of the number of projections). B2.3.2.4.4 Amount of Computer Data By setting up a scan and selecting the projection matrix, the amount of data produced in a scan should not be ignored. Independent of the required resolution, the reconstructed volume would be somehow related to the projection size. For example, an object scanned with a smaller matrix would be likely to be reconstructed in a smaller volume as well. For our Case 1 the volume data would lie somehow at 10243 and at 5123 voxels for Case 2. Since tomographic images are stored as integer values (Hounsfield unit [HU] numbers range from –1000 to a few thousand), the amount of computer data for the volumes would total to ~2 gigabyte for Case 1 and 256 megabyte for Case 2. Not every software and computer can cope with such large data sets.
B2.3.2.5 Summary Micro-CT is a modern, nondestructive tool for investigation of small samples, delivering 3D views of their anatomy. Potential applications range from in vivo medical to in vitro and ex vivo biological to technical and industrial. Experience gained in clinical CT can be transferred to micro-CT. However, specific features and properties of micro-CT samples may require selection of different scan settings in clinical CT in order to optimize image quality, including the following: r Selection of the proper x-ray beam energy (i.e., tube high voltage): The stronger the absorption of the specimen the higher must be the beam energy and high voltage. r Spatial resolution, which must be consistent with the sample size r Adjustment of the volume of measurement to the sample size
80
The Virtopsy Approach
In this brief introduction to micro-CT several issues have not been addressed at all, such as the following: r Verification of scanner performance and quality assurance [11,12] r Measurement and validation of the components’ parameters (e.g., x-ray tube focus size in the micrometer scale [14]) r Scanner mechanical alignment in a micro-scale [15,27] r Problems specific to ultrahigh spatial resolution (scanner long-term stability) [1] r Artifacts induced through insufficient temporal resolution (e.g., motion artifacts in in vivo scanning) These problems lie beyond the scope of this chapter, and interested readers are advised to look in the literature for further information. Also, the future of micro-CT has not yet been discussed. The common trend is to build faster scanners with better temporal and spatial resolution. Another interesting approach is to equip scanners with two tubes and two detectors to speed up acquisition but also to enable dual energy scanning (known from the clinical CT field). Multimodality imaging (i.e., combination of x-ray micro-CT with micro-PET/SPECT [positron emission tomography/single photon emission computed tomography] and micro-MR) is another example of the development in this field. All these aspects have not been included in this review. The intention was to get the reader acquainted with the basic principles of the micro-CT technique. We do hope to have conveyed the following message: r Spatial resolution and specimen size depend on each other: High resolution can be achieved only in small samples. A good rule of thumb is that best spatial resolution ≈ sample diameter/500…1000 (depending on the detector image matrix size in pixels). r Voxel size, at which the sample is being scanned, is not equal to resolution. A good rule of thumb is that spatial resolution ≥ (at least) twice the voxel size. r The sky is the limit, but the underlying physics cannot be fooled: r In vivo applications: There will always be a compromise among radiation dose, spatial resolution, and noise. r In vitro/NDT applications: The user must choose among scan time, spatial resolution, noise, and specimen size.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12. 13. 14.
15.
16.
17.
18.
B2.3.2.6 References 1. Haddad W.S., McNulty I., Trebes J.E., et al. 1994. Ultrahighresolution x-ray tomography. Science 266: 1213–1215. 2. Engelke K., Karolczak M., Lutz A., et al. 1999. High spatial resolution 3D x-ray cone-beam microtomography.
© 2009 by Taylor & Francis Group, LLC
19.
Radiology 213(P): 414, Annual RSNA Congress, November 1999, Chicago, USA. Flynn M.J., Hames S.M., Reimann D.A., et al. 1994. Microfocus x-ray source for 3D microtomography. Nuclear Instruments and Methods in Physics Research. 353: 312–315. Badea C.T., Fubara B., Hedlund L.W. et al. 2005. 4-D micro-CT of the mouse heart. Molecular Imaging 4: 110–116. Beuck L., Vertino A., Stepina E., et al. 2007. Skeletal response of Lophelia pertusa (Scleractinia) to bioeroding sponge infestation visualised with micro-computed tomography. Vol. 53, pp. 157 176. Berlin/Heidelberg Springer. Cnudde V., Cnudde J.P., C. Dupuis C., et al. 2004. X-ray micro-CT used for the localization of water repellents and consolidants inside natural building stones. Materials Characterization 53: 259–271. Karolczak M., Kachelrieß M., Ott O., et al. 2005. A highspeed micro-CT scanner with rotating gantry for in-vivo animal scanning. Biomedizinische Technik. 50 (Suppl. Vol. 1, part 1): 756–757. Engelke K., Karolczak M., Lutz A., et al. 1999. Mikro-CT. Technologie und Applikationen zur Erfassung von Knochenarchitektur. Der Radiologe 39: 203–212. Graeff W. and Engelke K. 1991. Microradiography and microtomography. In: Handbook on Synchrotron Radiation, vol. 4, pp. 361-405, North Holland: Elsevier. Boone J.M., Velazquez O., and Cherry S.R. 2004. Smallanimal x-ray dose from Micro-CT. Molecular Imaging 3: 149–158. Kalender W.A., Durkee B., Langner O., et al. 2005. Comparative evaluation: acceptance testing and constancy testing for micro-CT scanners. Biomedizinische Technik. 50 (Suppl vol. 1, part 1): 1192–1193. Kalender W.A. 2005. Computed Tomography, 2d ed. Erlangen: Publicis Verlag. Spanne P. 1989. X-ray energy optimization in computed microtomography. Phys. Med. Biol. 34: 679–690. Taubenreuther U., Engelke K., Riedel T., et al. 2002. Measurement of x-ray tube focal spot sizes and spatial resolution of MCT systems using thin wires. Radiology 209(P): 400, Annual RSNA Congress, November 2002, Chicago, USA. Karolczak M., Schaller S., Engelke K., et al. 2001. Implementation of a cone-beam reconstruction algorithm for the single circle source orbit with embedded misalignment correction using homogenous coordinates. Med. Phys. 28: 250–269. De Clerck N.M., Meurerens K., Weiler H., et al. 2004. High resolution x-ray microtomography for the detection of lung tumors in living mice. Neoplasia 6: 374–379. Engelke K., Wachsmuth L., Taubenreuther U., et al. 2001. High resolution in vitro MCT of osteoarthritis in a mouse model. In: High Resolution Imaging in Small Animals: Instrumentation, Applications and Animal Handling. Rockville MD. McErlain D.D., Chem R.K., Bohay R.N., et al. 2004. Microcomputed tomography of a 500-year-old tooth: technical note. Can. Assoc. Radiol. J 55: 242–245. Ford N.L., Thornton M.M., and Holdsworth D.W. 2003. Fundamental image quality limits for microcomputed tomography in small animals. Med. Phys. 30: 2869–2877.
Internal Body Documentation
20. Fuchs T. and Kalender W.A. 2003. On the correlation of pixel noise, spatial resolution and dose in computed tomography: theoretical prediction and verification by simulation and measurement. Physica Medica 19: 153–163. 21. van Geet M., Swennen R., and Wevers M. 2000. Quantitative analysis of reservoir rocks by microfocus x-ray computerised tomography. Sedimentary Geology 132: 25–36. 22. Grabherr S., Hess A., Karolczak M., et al. 2007. Blood-vessel visualization using Angiofil® and micro-computed tomography—a feasibility study, submitted for publication. 23. Harpen M.D. 1999. A simple theorem relating noise and patient dose in computed tomography. Med. Phys. 26: 2231–2234. 24. Holdsworth D.W. and Thornton M.M. 2002. Micro-CT in small animal and specimen imaging. Trends Biotechnol. 20, no. 8 (Suppl): 34–36. 25. Van Kaick G. and Delorme S. 2005. Computed tomography in various fields outside medicine. Eur. Radiol. Suppl 15 (suppl 4): D74–D81. 26. Karolczak M., Seibert U., Lutz A., et al. 2000. Cone-beam spiral CT: First experimental results. Radiology 217(P): 404 (Annual RSNA Congress, November 2000, Chicago, USA). 27. Karolczak M., Taubenreuther U., Lutz A., et al. 2001. Practical approach to misalignment correction in a singlecircle orbit cone-beam tomography, 3D. Paper presented at the Sixth International Meeting on Fully Three-Dimensional Image Reconstruction in Radiology and Nuclear Medicine, October 30–November 2, Pacific Grove, CA. 28. Krempien R., Huber P., Treiber M., et al. 2000. Combination of irradiation and bisphosphononates in the therapy of bone metastases: an experimental study. Paper presented at the American Society for Therapeutic Radiology and Oncology meeting, July. 29. Müller R., Van Campenhout H., Van Damme B., et al. 1998. Morphometric analysis of human bone biopsies: a quantitative structural comparison of histological sections and micro-computed tomography. Bone 23: 59–66. 30. Paulus M.J., Gleason S.S., Sari-Saraf H., et al. 2000. Highresolution x-ray CT screening of mutant mouse models. SPIE 3921: 270–279. 31. Reimann D.A., Hames S.M., and Flynn M.J. 1995 A flexible laboratory system for 3D x-ray micro-tomography of 3-50 mm specimens. 3D microscopy: image acquisition and processing. SPIE 2412-26. 32. Sasov A. 1987. Microtomography. J. of Microscopy 147: 189–192. 33. Stepina E., Fuchs T., Engelke K., et al. 2003. Dose determination for micro-CT in small animals. In: Medizinische Physik, ed. W. Semmler and L. Schad. Heidelberg: Deutsche Gesellschaft für Medizinische Physik. 34. Thali M.J., Taubenreuther U., Braun M., et al. 2001. Micro-CT and forensic pathology. Rechtsmedizin 11: 192. 35. Thali M.J., Taubenreuther U., Braun M., et al. 2002. Micro-CT: An advantageous tool for forensic pathology. ECR p. 163. 36. Thali M.J., Taubenreuther U., Karolczak M., et al. 2003. Forensic microradiology: micro-computed tomography (Micro-CT) and analysis of patterned injuries inside of bone. J. Forensic Sciences 48: 1336–1342. 37. Wachsmuth L. and Engelke K. 2004. High resolution imaging of osteoarthritis using microcomputed tomography. In: Methods in Molecular Medicine, Ed. F. De Ceuninck and M. Sabatini, 231–248. Totowa, NJ: Humana Press.
© 2009 by Taylor & Francis Group, LLC
81
B2.4 MAGNETIC RESONANCE IMAGING B2.4.1 BASICS OF MRI AND MR-SPECTROSCOPY Chris Boesch Magnetic resonance imaging (MRI) is well known in the medical community as an established radiological tool. Everyone has seen supreme images of the human body, in particular of soft tissue. However, MRI is just one part of nuclear magnetic resonance (NMR), as this method is called in basic science. The tremendous morphological information of MR images sometimes conceals the fact that MR signals in general contain much more information, especially on processes on the molecular level. This includes the chemical analysis of tissue using MR spectroscopy of hydrogen or other nuclei, in particular carbon and phosphorus. NMR is successfully used in physics, chemistry, and biology to explore and characterize chemical reactions, molecular conformations, biochemical pathways, solid state material, and many other applications that elucidate otherwise invisible characteristics of matter and tissue. This introduction shall illustrate how one physical effect—the ability of nuclei to absorb and emit radiofrequency (RF) if placed in a strong magnetic field—can lead to various applications in basic science and medicine. The range of applications is enormous and reaches from macroscopic images to the molecular level, where proteins and other molecules are characterized with the help of NMR. It is also important to recognize the enormous versatility of NMR and MRI as compared with other diagnostic modalities such as computer tomography (CT) or ultrasound (US). While CT and US are invaluable methods in radiological diagnostics, both modalities measure primarily a single physical characteristic of tissue, either permeability or density. MRI measures proton density, diffusion, oxygenation, motion, chemical environment, and many other parameters of tissue or in solutions. This makes the versatility of MR and NMR unique but also complicates their application. B2.4.1.1 Short History of NMR and MRI It is more than just historical interest that relates biophysical NMR and biomedical MRI. Since methods that have already been used in NMR can be transferred to MRI in one way or the other, NMR represents a pool of ideas for future MRI applications. The historical development of NMR and MRI has been described in detail in textbooks and review articles [1,2]. A particularly impressive historical fact is the large number of Nobel Prizes that have been awarded directly for NMRrelated work or for discoveries in related fields [3]. Around 1920 to 1930, when the physics of nuclei was studied, the angular momentum of nuclei (the “spin”) was discovered and the famous experiments in Otto Stern’s laboratory revealed observations that were only explainable in the newly developed quantum physics by a combination of angular momentum and magnetic moment of the nuclei and electrons. Stern was awarded the Nobel Prize in physics in 1943.
82
The Virtopsy Approach
Isidor Isaac Rabi was one of the scientists who aimed at an exact measurement of the nuclear magnetic moment. His successful experiments with oscillating fields led to Rabi being awarded the Nobel Prize in physics in 1944. With radiofrequency components that were improved by the development of radar during World War II, two groups independently conducted experiments on NMR in solids. The first group consisted of Edward Purcell, Henry Torrey, and Robert Pound, and the other group was made up of Felix Bloch, William Hansen, and Martin Packard. While it was not immediately clear that both groups described the same effect, Bloch and Purcell received the 1952 Nobel Prize in physics. Up to that point in time, the NMR effect was exclusively seen as a measure to determine nuclear properties, and nobody was able to foresee that this effect would revolutionize physics, chemistry, biology, and medicine. Around 1950, the observation that the chemical environment within molecules affected the NMR signal was a first step toward chemistry. This effect was called the “chemical shift” and represented the basis for an unparalleled development of NMR as an analytical method in chemistry and biophysics. The term high-resolution NMR was really earned when the field strength of the magnets used for NMR was increased and when Anderson and Ernst developed the so-called Fourier NMR technique, which was then further developed by Ernst’s group in multiple dimensions. Richard Ernst was awarded the 1991 Nobel Prize in chemistry for these developments. Kurt Wüthrich applied high-resolution NMR to study the structure of biological macromolecules and received the 2002 Nobel Prize in chemistry. When Paul Lauterbur and Sir Peter Mansfield in the early seventies published a way to encode the NMR signal spatially, the first steps for medical imaging were laid. Consequently, they were awarded the Nobel Prize in medicine in 2003. For the application of NMR in medicine, it was agreed that the “N” in “NMR” be omitted in order to
distinguish “MR” from nuclear medicine since MR leaves nuclei and electrons intact and thus creates no ionizing radiation. When one looks at the enormous number of Nobel prizes that are connected with the development of NMR and MRI, one can estimate the many steps that were made before the methods were as successful as they are today. MRI and NMR still develop with incredible speed, and the diversification in particular of MRI is such that even MR experts can’t cover more than just a limited part of medical MRI. B2.4.1.2 The Basics of the NMR Effect It is surprising that NMR and MRI are based on the very same physical effect, nuclear magnetic resonance. This effect occurs when certain nuclei—typically isotopes with an odd number of nucleons—are placed in a strong magnetic field. In the following, it shall be explained how this effect can be used to investigate tissue in situ noninvasively, how the spatial distribution can form images, and how the chemical environment can modulate the signals such that molecules can be studied in pure solutions and in situ. A large number of books and review articles deal with the basic principles and applications of NMR and MRI on various educational levels (Figure B2.4.1.1) [2,4–27]. A nucleus can have two attributes that are the basis for the NMR effect: a magnetic moment and an angular momentum (“spin”). These attributes can also be found in a rotating bar magnet (Figure B2.4.1.1) and lead to rather unique behavior. The angular momentum tries to keep the direction of the rotating bar magnet unchanged while the magnet is attracted by an external magnetic field. These diverging forces lead to a compromise: The bar magnet begins to precess (i.e., it neither aligns with the applied external field nor keeps on the original axis). The basis of NMR and MRI is the fact that the frequency of the resulting precession is exactly proportional to the applied magnetic field (Figure B2.4.1.2). In other words, if
Angular momentum “spin” Nucleus
1
H,13C P, ...
31
Precession
Induction, radio waves
D
E
Magnetic moment
A
B
C
FIGURE B2.4.1.1 This illustrates the basis for nuclear magnetic resonance in a classical model. (A) Nuclei as listed in Table B2.4.1.1 (e.g., 1H, 13C, or 31P) show two properties that lead to the nuclear magnetic resonance effect. (B) Nuclei possess an angular momentum (“spin”) comparable to a spinning top and a magnetic moment similar to a bar magnet. (C) An external magnetic field attracts the magnetic moment such that it tends to align along the external field; however, the angular momentum tries to keep the orientation of the axis unchanged. (D) Just like a spinning top on a table, a nucleus begins a precession, which is actually a compromise between the requirements of angular momentum and magnetic moment. (E) If this precession of many nuclei can be synchronized in a macroscopic sample, it can be detected since the resulting magnetization induces a voltage in a coil (a radiofrequency antenna).
© 2009 by Taylor & Francis Group, LLC
Internal Body Documentation
83
! between magnetic field can be varied by: and radiofrequency depending on nucleus (1H, 13C, 31P....)
FIGURE B2.4.1.2 The gyromagnetic ratio describes that the precession frequency (see Figure B2.4.1.1) is proportional to the magnetic field strength; that is, when the magnetic field strength changes, the detected radiofrequency follows the changes in an unambiguous manner and can be used to determine the magnetic field at the place of the observed nuclei. This effect can be used to spatially encode the signals (Figure B2.4.1.3) or to detect the chemical environment in molecules (Figure B2.4.1.4).
we are able to detect the precession frequency, spins report the exact magnetic field strength on an atomic level. Two questions come up: (1) How can we detect the precession frequency, and (2) how can we use this information to generate images and spectra? To the first question, a bar magnet in front of a coil induces an alternating voltage—such as in a dynamo. If spins are brought into an applied magnetic field, the precession of the spins is uncoordinated (“they are not in phase”). Using radiofrequency pulses, it is now possible to bring the spins into a coordinated precession. Like in a water ballet, the synchronized spins then are able to induce a voltage in a coil—that is, the spins emit an electromagnetic wave that can be detected outside of the body (Figure B2.4.1.1). Since the human body can be penetrated by electromagnetic waves of some Megahertz, one can detect signals from spins that are located deep inside the body. To the second question, if we are able to detect the signals of the spins as previously described, how can we use the fact that the frequency of these waves exactly tells us the magnetic field strength in the body or in a solution? To understand the enormous consequences of this fact, we need to look at possible mechanisms that may alter the magnetic field strength (Figure B2.4.1.2). One possible mechanism is used by technically induced magnetic field gradients; that is, one can make the magnetic field strength spatially dependent. If the frequency of the radiowaves is proportional to the magnetic field strength, we can tell the spatial position of a spin in a spatially varying magnetic field (Figure B2.4.1.3). These “gradients” are the basis for the spatial encoding of MR signals and will be discussed following. Another mechanism occurs within molecules where the density of the electron clouds shield the external magnetic field and lead to chemical
© 2009 by Taylor & Francis Group, LLC
Single frequency
Frequency variations = projection
Signal Intensity
Frequency reveals magnetic field variations
Spatial variation of magnetic field strength (“gradient”)
Spatially homogenous magnetic field strength
Signal Intensity
can be detected in radiofrequency coils (antennas)
Sample (two vials)
Frequency
Frequency
FIGURE B2.4.1.3 The spatial encoding of magnetic resonance signals is based on the spatial variation of the magnetic field strength. If a sample, such as two vials as indicated in the figure, is placed in a spatially homogenous field, all spins experience the same magnetic field strength and resonate at one specific frequency (left column). If the magnetic field strength is varied linearly in one direction (if a so-called gradient is applied, right column), the spins no longer experience the same field strength but resonate at an increasing frequency, depending on their exact location. The result of such an acquisition would be a one-dimensional projection as known from computed tomography.
information that is encoded in the magnetic field strength (Figure B2.4.1.4). Both mechanisms are explained in more details in the following paragraphs. The gyromagnetic ratio (Figure B2.4.1.2) reveals an additional fact: The term I defines the proportionality between magnetic field strength and resonance frequency. Since I is very different for various nuclei, they resonate at frequencies that are easily technically distinguishable (Table B2.4.1.1). To summarize, differences of the resonance frequencies between different nuclei are on the order of several Megahertz; gradient-induced magnetic field variations induce changes of several Kilohertz; and the influence of the chemical environment in molecules vary the resonance frequency by several Hertz. This explains why it is easy to distinguish signals from different nuclei, robust to encode space with magnetic field gradients however, rather subtle to measure a spectrum
84
The Virtopsy Approach
External magnetic field
Weak electron cloud
Strong electron cloud
Unshielded
Shielded
Higher frequency
Lower frequency Frequency (ppm)
FIGURE B2.4.1.4 This shows the origin of the “chemical shift.” Electrons form covalent chemical bonds between the nuclei of a molecule; they are the “glue” of a chemical compound. The different nuclei contribute to this glue to a different extent; electronegative nuclei such as oxygen attract more electrons such that the neighboring nuclei have less. Carbon, on the other hand, shares the electrons with the hydrogen nuclei. Since the clouds of electrons shield the nuclei form the external magnetic field, the neighborhood of oxygen experiences less shielding by the electrons and, therefore, a stronger contribution from the external magnetic field. In turn, hydrogen nuclei (“protons”) bound to carbon are shielded by the equally distributed electron clouds and, therefore, experience a weaker magnetic field. According to the gyromagnetic ratio (Figure B2.4.1.2), the stronger–weaker magnetic field turns into a higher– lower radiofrequency that can be detected.
of resonance frequencies that originate from the same type of nucleus in a molecule. B2.4.1.3 Gradients Used for Spatial Encoding in Imaging and Volume Selected Spectroscopy: Image Formation Based on the “gyromagnetic ratio,” we know that the resonance frequency of nuclei is proportional to the applied magnetic field. If we now alter the magnetic field in space, the resonance frequency changes with the position in space; that is, the frequency can be used to label the space (Figure B2.4.1.3). One can use a piano as an analog: If we hear the sound from a piano, we know the position of the piano player
© 2009 by Taylor & Francis Group, LLC
(or to be more precise—of the finger that played the single tone). Similarly, the frequency of the radiofrequency signal from a nucleus tells us the position—at least in the single direction where the magnetic field is varying. This variation of the magnetic field strength is called magnetic field gradient or, briefly, gradient. Gradients can be used during excitation or during acquisition to encode the space. During acquisition, the MR sequence applies pulses with a narrow bandwidth (distribution of frequencies that contribute to the pulse) to bring the nuclei to a synchronous precession. According to the gyromagnetic ratio condition, just nuclei within a slice fulfill the resonance condition and get into such a synchronous precession that their response can be observed (i.e., the radiofrequency pulse selected a slice), which will produce signals in the following MR sequence. During acquisition, another mechanism is used for spatial encoding. If the excited spins emit radiofrequency signals while they are located in a magnetic field gradient, the distribution of resonance frequencies will produce a projection; that is, signals from nuclei with the same position along the gradient will add together (Figure B2.4.1.3). This projection was used in the beginning to generate images similar to computed tomography; that is, projections from different angles were used to generate an image by back-projection algorithm. Following Ernst’s development of two-dimensional NMR, he also suggested use of this Fourier method for image generation. While one spatial dimension is encoded during acquisition with the projection along the so-called read gradient, the other dimension is encoded stepwise between excitation and acquisition with a gradient along that dimension. These phase steps can be seen as a stepwise sampling that leads to the same result as the read gradient during acquisition. While this Fourier imaging is rather abstract and difficult to understand for beginners, it is in fact one of the basis for the enormous versatility of MRI. Magnetic resonance spectroscopy (MRS) is a combination of NMR and MRI that reveals the chemical information (NMR) from a defined region in the body (MRI). Since the chemical information is sampled during acquisition, no additional spatial encoding is done during acquisition. For spatial encoding outside the acquisition window, in principle two techniques can be used: either phase gradients in a spatial dimension (chemical shift imaging) or volume selection during excitiation (single voxel spectra). B2.4.1.4 Chemical Information We have seen that the gyromagnetic ratio relates magnetic field strength and resonance frequency. While it is clear that a spatial variation of the magnetic field strength leads to changes of the frequency, there is another mechanism that can be observed in molecules. Even if molecules are placed in a perfectly homogenous magnetic field, the chemical environment influences the magnetic field strength within the molecule. Since the different nuclei in a molecule attract the surrounding electrons to varying degrees, the
Internal Body Documentation
85
TABLE B2.4.1.1 Selected Nuclei That Can be Used to Observe an NMR Signal
$$"
$!
"#
+
&
K\GURJHQ SURWRQV FDUERQ
&
FDUERQ
) IOXRULQH 1D VRGLXP 3 SKRVSKRUXV
! $& #" '
!# ""#%#& #"##( !#%#
#$! &#!#" $ $#""$
QR105 VLJQDO
,#!)* + LQPHWDEROLWHV DGLSRVHWLVVXH
&%! ""#%#& !#%!! #$&
DGLSRVHWLVVXH LQPHWDEROLWHV
Note: MRI uses mainly the signals from protons in water while MRS observes the signal from nuclei with considerable concentrations in tissue such as 1hydrogen (protons), 13carbon, 31phosphorus, and a few others. “Typical Concentrations of the Nuclei” and “Overall Sensitivity” are rough estimations for an illustration of the order of magnitude only. “Typical Overall Sensitivity” for nuclei and metabolites that are typically observed by in vivo MRS is drastically lower than the sensitivity of water in 1H-MRI as it is shown on the top line. 12Carbon is added to this list to illustrate the fact that 99% of the carbon in the human body is NMR invisible.
© 2009 by Taylor & Francis Group, LLC
Free induction decay
Echo
Time Fourier-
Amplitude
density of the electrons in molecules is spatially dependent (Figure B2.4.1.4). Since electrons tend to shield the nuclei from the applied external magnetic field, nuclei within a strong electron cloud experience less magnetic field strength than others. For example, neighbors of an oxygen nucleus in a molecule have less electron density since oxygen attracts the electrons. Subsequently, the neighbors experience a stronger magnetic field—that is, the basis for the chemical information that can be reported by spins, leading to so-called (N) MR spectra of molecules. A single, one-dimensional spectrum is defined by the chemical shift axis, which is expressed in parts per million (ppm) of the applied magnetic field (Figure B2.4.1.5). This unit is well chosen since the chemical shift scales with the magnetic field strength and, therefore, the relative position of a resonance measured in ppm is independent of the magnetic field applied. The signal intensity is proportional to the amount of nuclei that contribute to the resonance; in other words, while the x-axis determines the chemical species, the y-position gives the amount of this species. However, other factors such as relaxation times and acquisition parameters also influence the signal intensity and have to be controlled if quantitative measurements are done. An inherent problem of NMR is the fact that the signal reception in the antenna and the signal amplification introduce arbitrary factors—that is, the signal needs some calibration before it can be used in absolute terms. This chemical information can now be used in two different ways, either in pure solutions of one or just very few substances or in a mixture of substances in vitro or in vivo. High-resolution NMR uses relatively well-defined, pure solutions and aims at a characterization of the molecules in that solution. This can be a known solution where the structure (“conformation”) of a specific molecule or group of molecules is studied. As an alternative, NMR can be used to follow
Coupling constant
Transformation
Linewidth
Coupling constant
Area Reference
6
5
4 3 Chemical shift
2
1
0 ppm
FIGURE B2.4.1.5 This defines parts of a (one-dimensional) magnetic resonance spectrum. The data acquisition, generated by the transversal magnetization that induces voltages in the receiver coil (antenna), is a signal that decays with time (“free induction decay”). Under specific circumstances, part of the decaying signal can be recovered, forming an echo. Following a so-called Fourier transformation, the different frequencies in free induction decay or echo can be separated into a spectrum as shown in the figure. Signal area, line width, and amplitude are related to the amount of nuclei that contribute to the resonance while the position on the chemical shift axis (given in parts per million from a reference substance) depends on the chemical nature of the molecules.
86
The Virtopsy Approach
reactions within solutions aiming at a characterization of the products. Either high-resolution NMR or in vivo MRS can be used to study mixtures of molecules in a solution. Highresolution NMR of such a mixture has the advantage that lots of resonances can be observed in parallel without destroying the sample. In vivo, the resolution of the spectra is limited due to the distortion of the magnetic field homogeneity by the introduced body and also by the interaction of molecules with solid parts of the tissue. Due to the overlapping signals in a mixture—in particular in vivo with limited resolution— the separation of the different signals is an important part and is done by various techniques of data fitting.
B2.4.1.5 Relaxation Times and Other Contrast Mechanisms It has been shown above that the MR signal contains information about the spatial origin (leading to images) and about the chemical environment (leading to spectra). In addition to this information, the MR signal can be varied by various contrast mechanisms, when contrast is used in a very general sense because various properties can be shown as signal intensity in an image, including speed, diffusion and other properties of tissue. The most traditional contrast mechanisms are based on the differences of the relaxation times in tissues. Relaxation times describe the way how magnetization behaves after excitation (Figure B2.4.1.6). The transversal relaxation time T2 defines the time how long the signal can be detected; that is, if a acquisition sequence waits a relatively long time before the signal is detected, tissue with short T2 will show a weaker signal than tissue with long T2. The longitudinal relaxation time T1 describes the return of the magnetization to the resting level; that is, sequences with a short time between the phase steps as previously mentioned will not give the magnetization enough time in tissue with long T1, leading to a weaker signal than from tissue with short T1. With appropriate sequences, the contrast of MR images can be varied considerably (Figure B2.4.1.7). Contrast agents (based on lanthanides or on ultrasmall ferromagnetic particles) vary the relaxation times and, therefore, can change signal intensity in an image. Various mechanisms of the tissue can be studied based on contrast agents, including flow of contrast agents in vessels for angiography, detection of impaired blood-brain barrier, inflow into tumors, uptake by macrophages, or the specific detection of antigens by contrast agents linked to antibodies as it is currently developed. Macroscopic flow can be detected and quantified based on the fact that excited nuclei can be detected after a certain time at a different place in space or that motion in a magnetic field gradient changes the resonance frequency. The latter effect can also be used to quantify microscopic motion (i.e., diffusion). Since diffusion tells us a lot about the microscopic structure of tissue, it is increasingly used to probe tissue changes during malignant transformation or with cellular
© 2009 by Taylor & Francis Group, LLC
Recovery of the longitudinal magnetization (available for next acquisition)
Longitudinal relaxation time T1 Time
Transversal relaxation time T2
Decay of the observable transversal magnetization (available for current acquition)
Time
FIGURE B2.4.1.6 This shows decay of the transversal and recovery of the longitudinal magnetization. For detection, the magnetization is tilted into the transversal plane (red arrow). This transversal magnetization decays with a time constant T2 while the longitudinal magnetization (blue arrow along the external magnetic field) recovers with a time constant T1. T1 and T2 are called relaxation times, describing how fast the detectable transversal signal exists and when the longitudinal magnetization is recovered for a next acquisition. Since different tissues do not have identical relaxation behavior, these mechanisms can be used to introduce image contrast between different tissues.
changes after stroke. In addition, the direction of anisotropic diffusion can be detected using MRI and can be used to probe the orientation of axons, subsequently leading to fiber tracking (i.e., the spatial course of nerve fibers). Since hemoglobin in blood changes its magnetic properties if it is either oxygenated or deoxygenated, it can be used to probe the oxygenation level of tissue. The most popular technique is blood-oxygen-level-dependent (BOLD) MRI, which can be used to follow brain activity, leading to images with color-encoded regions where the subject activated the brain. Many other contrast mechanisms are beyond the topic of this chapter, such as magnetization transfer, temperature measurements, hyperpolarization, and other properties of tissue. These contrast mechanisms are mentioned here just to illustrate the enormous versatility of MR. B2.4.1.6 Conclusion and Outlook In order to anticipate the potential of MRI and MRS for future development, it is also crucial to see these methods in the context with high-resolution NMR. Only this view sheds light on the enormous versatility and applications that have not been used so far. High-resolution NMR introduced methods that could be a source for many new applications in situ and also represents a chance for forensic medicine in particular for the investigation of tissue components with unknown
Internal Body Documentation
87
T1
T1
T1
T2
T2
T2
B
C
A
FIGURE B2.4.1.7 This example displays how the relaxation times can be used to introduce image contrast. (A) Proton density weighting: The acquisition is started soon after the excitation without losing too much signal due to T2 processes, and the next acquisition waits a sufficient time such that the longitudinal magnetization is fully recovered. That reduces the influence of T1 and T2, resulting in an image that shows more or less the amount of nuclei that contribute to the signal (“proton density”). (B) T 2 weighted image: The acquisitions are sufficiently apart such that the magnetization can recover and T1 has no influence on the image signal. Since the acquisition starts after a certain time, the T2 decay starts to reduce the signal intensity, at least in tissue with short T2 times. Tissues with longer T2 times (e.g., liquor) remain to give a strong signal. (C) T1 weighted contrast: If the acquisition is started soon after excitation, the influence of T2 is minimal. However, if the subsequent acquisition starts soon after the last one, the longitudinal magnetization cannot fully recover, at least in tissue with a long T2 time.
composition. The acquisition of an NMR spectrum typically does not require enormous preparation of the sample; it is nondestructive; and it reveals a wealth of parallel information about many metabolites. Optical methods are extremely sensitive in comparison with NMR; however, the sample needs special treatment, and, typically, it is no longer usable after the examination. In addition, chemical and optical methods are very specific for a specific metabolite and may not be suited to search for unknown products in the tissue. Imaging in forensic sciences will show various developments. A crucial issue will be standardized and comprehensive studies about all possible representations of forensic findings in MRI. Reduced temperature, tissue changes according to death, lack of perfusion, or signs of putrification change the contrast behavior of tissue in MRI. Other sequences, or at least adapted acquisition parameters, may help to cope with these changes. New developments with whole MR systems (in particular, radiofrequency coils that cover the whole body or moving tables) will help to use MR as a screening tool. However, it is important that MRI is not seen as CT where a few parameters can be used to adjust the parameters (Figure B2.4.1.7). Even for a quick screening, one should use the best optimized sequences to cover the major pathologies seen during autopsy. For example, searching a cardiac infarction or a comprehensive documentation of gunshot channels requires the application of different and specialized MRI-sequences. A forensic
© 2009 by Taylor & Francis Group, LLC
screening method based on MRI shall use recent technical developments such as arrays of radiofrequency coil and moving tables; however, an agreement on optimal acquisition methods is equally important if all, or at least the most important, findings of an autopsy should be covered. If the versatility of high-resolution NMR, MRI, and MRS is acknowledged, these methods can play a mandatory role in forensic medicine of the future. B2.4.1.7 References 1. Becker, E.D., Fisk, C., and Khetrapal, C.L., The development of NMR, in Encyclopedia of Nuclear Magnetic Resonance, Grant, D.M. and Harris, R.K., Eds., Wiley, Chichester, UK, 1–160, 1996. 2. Boesch, C., Molecular aspects of magnetic resonance imaging and spectroscopy, Mol. Aspects Med., 20, 185–318, 1999. 3. Boesch, C., Nobel prizes for nuclear magnetic resonance: 2003 and historical perspectives, J. Magn. Reson. Imaging, 20, 177–179, 2004. 4. Bigler, P., NMR Spectroscopy. Processing Strategies, 2d ed., Wiley-VCH, Weinheim, Germany, 2000. 5. Gadian, D.G., NMR and Its Applications to Living Systems, 2d ed., Oxford University Press, Oxford, 2002. 6. Boesch, C., Magnetic resonance spectroscopy: Basic principles, in Clinical Magnetic Resonance Imaging, 3d ed., Edelman, R.R. et al., Eds., Saunders, Elsevier, Philadelphia, 459–492, 2005.
88
The Virtopsy Approach
7. De Graaf, R.A., In Vivo NMR Spectroscopy: Principles and Techniques, John Wiley & Sons, Chichester, UK, 1999. 8. Mukherji, S.K., Clinical Applications of Magnetic Resonance Spectroscopy, John Wiley & Sons, New York, 1998. 9. Freeman, R., Magnetic Resonance in Chemistry and Medicine, Oxford Press, Oxford, 2003. 10. Salibi, N. and Brown, M.A., Clinical MR Spectroscopy: First Principles, Wiley, New York, 1997. 11. Horowitz, A.L., MRI Physics for Radiologists: A Visual Approach, 3d ed., Springer-Verlag, New York, 1995. 12. Gillies, R.J., NMR in Physiology and Biomedicine, Academic Press, San Diego, 1994. 13. Hendrick, R.E., Russ, P.D., and Simon, J.H., MRI: Principles and Artifacts, Raven Press, New York, 1993. 14. Stark, D.M. and Bradley, W.G., Magnetic Resonance Imaging, 2d ed., Mosby-YearBook, St. Louis, 1992. 15. Tofts, P., Quantitative MRI of the Brain: Measuring Changes Caused by Disease, John Wiley & Sons, Chichester, UK, 2003. 16. Nelson, J.H., Nuclear Magnetic Resonance Spectroscopy, Prentice Hall, Upper Saddle River, NJ, 2002. 17. Kuperman, V., Magnetic Resonance Imaging: Physical Principles and Applications, Academic Press, San Diego, 2000. 18. Sprawls, P., Magnetic Resonance Imaging: Principles, Methods, and Techniques, Medical Physics Publishing, Madison, WI, 2000. 19. Young, I.R., Grant, D.M., and Harris, R.K., Methods in Biomedical Magnetic Resonance Imaging and Spectroscopy, John Wiley & Sons, Chichester, UK, 2000. 20. Haacke, E.M., Brown, R.W., Thompson, M.R., and Venkatesan, R., Magnetic Resonance Imaging: Physical Principles and Sequence Design, John Wiley & Sons, New York, 1999. 21. Lufkin, R.B., The MRI Manual, 2d ed., Mosby, St. Louis, 1998. 22. Macomber, R.S., A Complete Introduction to Modern NMR Spectroscopy, Wiley-Interscience, Philadelphia, 1998. 23. Lambert, J.B. and Mazzola, E.P., Nuclear Magnetic Resonance Spectroscopy: An Introduction to Principles, Applications, and Experimental Methods, Prentice Hall, Upper Saddle River, NJ, 2003. 24. Levitt, M.H., Spin Dynamics: Basics of Nuclear Magnetic Resonance, John Wiley & Sons, New York, 2001. 25. Hore, P.J., Jones, J.A., and Wimperis, S., NMR: The Toolkit, Oxford University Press, Oxford, 2000. 26. Freeman, R., Spin Choreography: Basic Steps in High Resolution NMR, Oxford University Press, Oxford, 1998. 27. Edelman, R.R., Hesselink, J.R., and Zlatkin, M.B., MRI: Clinical Magnetic Resonance Imaging, 2d ed., Saunders, Philadelphia, 1996.
B2.4.2 VIRTUAL HISTOLOGY BY MAGNETIC RESONANCE MICROSCOPY Kimberlee Potter, William R. Oliver, and Michael J. Thali
the appropriate resonant frequency. The source of the emitted signal is not from the applied RF radiation itself, but from magnetic energy stored by the system when placed in an external magnetic field. Magnetic energy can be stored by nuclei with nonzero spin-angular momentum. In biological imaging, protons (mostly from water) are the most common source of signal because they are abundant in tissues and they have a high magnetic receptivity. The RF radiation emitted from different locations in the sample is spatially encoded by the superposition of magnetic field gradients on the external magnetic field. For in-depth coverage of the concepts just introduced, see Chapter B2.4.1 in this volume as well as several excellent books on the subject [1,2]. Imaging contrast depends on the total number of observed nuclei present in the sample and the rate at which the MRI signal decays with time following excitation. Signal decay is described by two characteristic relaxation times, T1 and T2. The T1 (or longitudinal) relaxation time measures the rate at which the spin system returns to equilibrium after RF excitation. T1 is accelerated if the spin system can lose packets of energy to energy-requiring processes at the resonant frequency. The T2 (or transverse) relaxation time describes the rate at which energy is lost by irreversible, entropic processes within the spin system. This process is accelerated by magnetic field perturbations, which be can very large if the magnetic dipoles are relatively fixed in orientation and position. Both T1 and T2 are very sensitive to tissue composition and therefore yield images with excellent soft-tissue contrast. Higher field strength magnets and very strong magnetic field gradients have facilitated the detection of the MRI signal from very small volume elements and have allowed images with high spatial resolution to be generated. These hardware advances have led to the development of a new branch of MRI called magnetic resonance microscopy (MRM) [1,3]. The principles of image generation are the same as for MRI, but the images that are generated have submillimeter resolution. The first MR microscope was realized in 1986 when it was used to image a single cell [4]. Today, MRM is widely exploited to study models of disease in laboratory animals [5,6], to generate three-dimensional anatomic and developmental atlases [7–9], and to provide rapid phenotyping of transgenic animals [10,11]. In this chapter, we intend to demonstrate how MRM can be used as an alternative to optical microscopy, given that MRM can provide true three-dimensional tissue information with less morphological artifacts associated with paraffin embedding and sectioning, and it has shorter data-acquisition times compared with the preparation of histological sections [12]. With its inherent advantages, MRM is a promising technique to study the nature and pattern of soft-tissue injuries in forensic medicine.
B2.4.2.1 Introduction
B2.4.2.2 Application of MRM to Wound Documentation
In magnetic resonance imaging (MRI), an image is constructed by spatially encoded radiofrequency (RF) radiation emitted by an object after the application of RF radiation at
The accurate assessment and objective documentation of a wound is of prime importance in forensic medicine. Typically, the forensic examination of a wound involves the
© 2009 by Taylor & Francis Group, LLC
Internal Body Documentation
89
[18,19]. These techniques give detailed information about changes to cell morphology in sections taken at the site of the electric entrance and exit wounds but provide little information about the extent of tissue damage in peripheral and deep tissues. MRI scans provide some information about vessel patency and muscle necrosis in survivors of electric shock, but the injury pattern is lost due to limited spatial resolution [20–23]. High-resolution images of human skin in vivo have been acquired previously with MRI scanners using specially designed surface coils and high magnetic field gradient coils [24,25]. These studies, however, focused on the structural and chemical changes in skin due to aging, drugs, and cosmetic products [26–28]. For the first time, MRM has been used to study the pattern of an electric injury in human skin. The zonal arrangement of tissue damage observed by MRM was consistent with literature reports [18], and sections treated with a variety of stains established the extent and severity of the tissue damage in the central, intermediate, and peripheral zones. Normal dermal tissue, in the peripheral zone (P), composed of large quantities of collagen and proteoglycans stained red-brown with Hinshaw-Pearse (Figure B2.4.2.1A), while the intermediate zone (I) revealed reduced staining intensity consistent with dermal edema. In the central zone (C), thermally damaged dermal tissue and the occlusions observed in the vessels of the superficial vascular plexus stained black with Hinshaw-Pearse, consistent with heat-induced changes [29].
visual inspection and photography of the injury pattern in combination with a forensic autopsy and histologic examination. In addition to conventional forensic techniques, modern radiographic methods, such as computed tomography (CT) and MRI, can be used prior to autopsy to assess the extent and severity of tissue damage in three dimensions. In many cases, however, the resolution of clinical scanners is not sufficient to answer questions relevant to forensic wound analysis. Emerging technologies with submillimeter resolution, such as MRM, may provide the requisite resolution for studying electric injuries whose severity, extent, and even presence can be difficult to establish [13]. The pathophysiology of electric trauma, involving thermal damage, electroporation, electrochemical effects, and, for high-energy arcs, blunt trauma, is incompletely understood [14–16]. The stigma varies widely depending on the strength and frequency of the electric field, the path of the current, and the histoarchitecture of the tissues [15–17]. Furthermore, electric trauma may be difficult to diagnose if characteristic skin lesions are absent. In the work presented, the microanatomy of an electric-injury pattern in human skin was characterized by MRM, and the extent and severity of tissue damage were correlated with gross and histologic findings for the wound site and the associated deep tissues. Various techniques have been applied to characterize the electric-injury pattern in skin, ranging from gross photographs to histology and scanning electron microscopy
A
C
C
I
P
B
D
FIGURE B2.4.2.1 (A) Hinshaw-Pearse stained histologic section showing thermally damaged tissues in the central (C), intermediate (I), and peripheral (P) zones of the electric-injury wound site. The red arrow indicates the carbonized area, and the black arrows indicate occluded blood vessels in the central zone. (B) Proton density image of the exit wound showing an injury pattern with a dark carbonized central zone (red arrow) with dark thombosed blood vessel (white arrow) and bright edematous intermediate zone. (C) T2-weighted image extracted from a 3D data set showing the course of the occluded vessel, the dark carbonized area, and the bright intermediate zone. Arrows indicate the hypothetical current path. (D) 3D volumetric image showing blood vessels (yellow) with occlusions (red) and an area of edema (green) segmented from the 3D MR data. The arrows denote the proposed path of the current. The 3D image was acquired with a RARE imaging sequence (TR/TE 2000/8 ms, NEX = 1, RARE = 8) at 150 Mm isotropic resolution.
© 2009 by Taylor & Francis Group, LLC
90
A representative water proton density map through the wound site in Figure B2.4.2.1B reveals a dark central zone, consistent with the reduced tissue hydration expected for heatcoagulated dermis, and a bright intermediate zone, consistent with the accumulation of interstitial fluid due to cellular edema and necrosis. The appearance of the exit wound was further analyzed by three-dimensional MRM. In T2-weighted images extracted from a 3D data set, the extent of severe thermal damage was seen as a reduction in signal intensity, while the extent of edema was seen as an enhancement of signal intensity. For example, the image shown in Figure B2.4.2.1C was extracted from a 3D data set but was oriented along the course of the superficial vascular plexus. There was a focal reduction in image intensity in the vessel lumen compared with other vessels present in the hypodermis. This corresponded to the vessel occlusions seen histologically. The occluded vessels and the area of edema were segmented interactively from the 3D MRM image and are rendered in Figure B2.4.2.1D to demonstrate the proposed path of the current along the vessels in three dimensions. This three-dimensional rendering of the vascular plexus and nearby edema supported the hypothesis that current traveled through the plexus until it arrived at a vessel that was in close proximity to the skin surface, where it arced to the ground, generating electrical and thermal damage in the tissues in its path. An advantage of MRM, which complements traditional histologic studies, is that the wound analysis is not limited to the underlying cutaneous tissue but includes the peripheral and deep tissues of the intact sample. Since all involved tissues can be examined at the same time in three dimensions, this type of analysis can be used to document the extent and severity of the damage sustained by cutaneous tissues. In the case presented here, MRM allowed us to classify and delineate, in three dimensions, regions with a unique appearance that correlated with histologically distinct tissue damage. This type of analysis can be valuable in the clinical management of these injuries, particularly in burn-depth estimation, as well as in the development of forensic strategies to discriminate this type of injury from other types of trauma to the skin.
B2.4.2.3 Mapping of Retinal Hemorrhage in Abusive Head Trauma Cases by MRM The existence and pathobiology of shaken baby syndrome (SBS) is a matter of some controversy in forensic medicine. As first described by John Caffey, SBS consists of retinal hemorrhage, subdural hemorrhage, and metaphyseal fractures of long bones in the absence of evidence of a blow to the head [30]. More commonly, the diagnosis of SBS is made on the basis of retinal hemorrhage, subdural hemorrhage, and diffuse axonal injury [31]. However, biomechanical models and other studies have been critical of the originally proposed mechanism for SBS, and many SBS cases have since been reclassified as abusive head trauma [32,33]. Multiple studies have demonstrated that retinal hemorrhage is a ubiquitous finding in severe abusive head trauma
© 2009 by Taylor & Francis Group, LLC
The Virtopsy Approach
and is common in catastrophic nonabusive trauma (e.g., deaths due to motor vehicle accidents and falls from great heights) [34]. Additionally, retinal hemorrhage has been documented in cases of other trauma, including cardiopulmonary resuscitation [35–37] or after short falls [38]. These findings have led to the questioning of retinal hemorrhage as a valid diagnostic criterion for abusive head trauma. Morphological reviews have described qualitative differences between retinal hemorrhage from abusive head trauma and hemorrhage associated with resuscitation or short falls [39]. In abusive head trauma, the hemorrhages are florid, deep, and distributed widely, while in nonabusive events such as resuscitation the hemorrhages are described as small, rare, and superficial. However, the significance of retinal hemorrhages in the diagnosis of abusive head trauma is under debate in part because of the subjective method of describing the lesions. A more quantitative approach to retinal hemorrhage may better elucidate its diagnostic significance. We propose to use MRM to quantify and map retinal hemorrhage. While technical issues remain, the method promises to provide a metric for the evaluation and grading of retinal hemorrhage that will aid in the characterization of this lesion. The advantages of the MRM method are that it provides a quantitative estimate of hemorrhage volume with direct spatial mapping to the eye and is nondestructive. Imaging protocols were designed to acquire 3D images of the eye with sufficient resolution to resolve the retinal surface of the eye and, with the appropriate contrast, to allow for the discrimination of hemorrhage from the fluid-like vitreous and the capillary-rich choroid layer. In the T1-weighted gradient echo image (Figure B2.4.2.2A), the connective tissue layers of the eye were readily discerned. The outermost layer, containing the conjunctiva (anterior aspect) and attached musculature, was of intermediate intensity, and the dense connective tissue of the sclera was dark. The subjacent choroid had the highest intensity, consistent with the presence of blood in this layer. In this particular specimen, from an SBS or abusive head trauma case, the retina appears relatively thick and dark (Figure B2.4.2.2A), owing to large florid hemorrhage throughout the retina extending along the optic nerve sheath. The retinal hemorrhage observed by MRM was consistent with that observed on microscopic examination of histologic sections taken through the eye (Figure B2.4.2.2B). The dark appearance of the hemorrhage in the MRM image was consistent with that of an old hemorrhage containing deposits of the breakdown products of hemoglobin [40,41]. A volumerendered 3D gradient echo image of the eye is presented in Figure B2.4.2.2C. The vitreous was removed from the data set (made black) by a gradient flood fill and a simple threshold. Intraretinal and subarachnoid hemorrhages, dark regions in the MRM data set, were segmented in a semiautomated manner using adaptive thresholding and rendered as a separate data set with a high opacity and colored red. To visualize hemorrhage in the context of the eye anatomy, the eye and hemorrhage images were merged in a 3D viewer in AVS/Express (Advanced Visual Systems, Waltham, MA).
Internal Body Documentation
A
C
91
B
D
FIGURE B2.4.2.2 (A) Two-dimensional MRM image of the left globe from an abusive head trauma case in an infant. Image was extracted from a three-dimensional data set acquired using a gradient echo pulse sequence (TR/TE 200/3 ms, NEX 2, AQ 7h) with a voxel resolution of (86 Mm)3. (B) Histologic section of the eyeball, stained with Hematoxylin-Eosin, showing the distribution of hemorrhage in the retina and around the optic nerve sheath. (C) Volumetric rendering of the 3D MRM data set of the eyeball with the hemorrhage in opaque red. (D) Cross-sectional view of the eyeball with a white sclera, yellow lens, and red hemorrhage covering most of the surface of the retina.
Regions of florid hemorrhage were consistent with those observed by macroscopic inspection (Figure B2.4.2.2D). The observed retinal detachment in histologic sections was artifactual and was introduced when the eye was removed from the skull and exacerbated during tissue dehydration for the paraffin embedding process. Histologic sections did, however, confirm that much of the dark areas observed by MRM corresponded to retinal hemorrhage, and MRM images reflected both the distribution and the timing resolution of the hemorrhage. There have been numerous studies on the appearance of hemorrhage in MRI images [40,41]. Based on these reports, we have determined that the age of the hemorrhage will greatly impact its appearance in the MRI image. Optimum contrast is best achieved if the hemorrhage contains hemosiderin and ferritin deposits, which make it distinguishable from the choroid layer. If the hemorrhage is fresh there might not be sufficient hemoglobin decomposition to provide contrast for hemorrhage visualization. Also, if the hemorrhage is mostly superficial, the resolution of our current imaging method may be insufficient to resolve the signal loss on the very thin retinal surface. To improve the sensitivity of our MR experiments to studies of retinal hemorrhage, imaging coils based on a phased-array design might be used to image the surface of the eye at the expense of the vitreous.
© 2009 by Taylor & Francis Group, LLC
B2.4.2.4 Future Prospects Many forensic specimens submitted for MRM analysis are formalin fixed. Unfortunately, formalin-fixed tissues tend to be less hydrated, with shorter T2 and longer T1 values compared with living tissues [42]. Given the lack of blood flow or tissue oxygenation, there is little contrast difference in formalin-fixed tissues, and the long T1 values can result in long image-acquisition times for diagnostic images. Imageacquisition times can be reduced if the specimens are perfused with [12], or immersed in [43], a fixative containing a suitable contrast agent that will reduce the T1 relaxation times of the tissues. More importantly, the differential uptake of the contrast agent by the different tissues within the specimen can vastly improve the resulting image contrast. Notably, MRM contrast agents do not appear to impact the outcome of histology staining [44]. Alternatively, image-acquisition times can be reduced by wrapping the specimen in multiple receiver coils, namely a phased-array coil, and detecting the MRM signal from each coil simultaneously rather than sequentially in time. This approach reduces the total imaging time and improves the signal reception because of the proximity of the coil array to the sample [45,46]. With this type of coil arrangement the surface of the specimen is detected with high sensitivity, which is especially useful for forensic studies of the eye and the skin. Finally, it is conceivable that a phased-array coil might be used to produce a three-dimensional excitation with simultaneous transmission of RF pulses to the multicoil array such that a unique region of interest can be selectively excited and thereby imaged at much higher spatial resolution [47,48]. Using this approach, termed parallel transmission, it is foreseeable that routine clinical MRI machines will be in a position to perform high-resolution scans on regions selected from whole-body MRI scans. This innovation promises to greatly increase the utility of the MRI scanner in the performance of a forensic autopsy. B2.4.2.5 References 1. Callaghan, P. T. 1991. Principles of Nuclear Magnetic Resonance Microscopy. Oxford: Oxford University Press. 2. Morris, P. G. 1986. Nuclear Magnetic Resonance Imaging in Medicine and Biology. Oxford: Clarendon Press. 3. Kuhn, W. 1990. NMR-microscopy—fundamentals, limits and possible applications. Angew Chem Int Ed Engl 29:1–112. 4. Aguayo, J. B., S. J. Blackband, J. Schoeniger, M. A. Mattingly, and M. Hintermann. 1986. Nuclear magnetic resonance imaging of a single cell. Nature 322:190–91. 5. Benveniste, H. and S. Blackband. 2002. MR microscopy and high resolution small animal MRI: applications in neuroscience research. Prog Neurobiol 67:393–420. 6. Epstein, F. H. 2007. MR in mouse models of cardiac disease. NMR Biomed 20:238–55. 7. Benveniste, H., K. Kim, L. Zhang, and G. A. Johnson. 2000. Magnetic resonance microscopy of the C57BL mouse brain. Neuroimage 11:601–11. 8. Dhenain, M., S. W. Ruffins, and R. E. Jacobs. 2001. Threedimensional digital mouse atlas using high-resolution MRI. Dev Biol 232:458–70.
92
The Virtopsy Approach
9. Ruffins, S. W., M. Martin, L. Keough, et al. 2007. Digital three-dimensional atlas of quail development using highresolution MRI. Scientific World Journal 7:592–604. 10. Johnson, G. A., G. P. Cofer, S. L. Gewalt, and L. W. Hedlund. 2002. Morphologic phenotyping with MR microscopy: the visible mouse. Radiology 222:789–93. 11. Nieman, B. J., N. A. Bock, J. Bishop, et al. 2005. Magnetic resonance imaging for detection and analysis of mouse phenotypes. NMR Biomed 18:447–68. 12. Johnson, G. A., H. Benveniste, R. D. Black, L. W. Hedlund, R. R. Maronpot, and B. R. Smith. 1993. Histology by magnetic resonance microscopy. Magnetic Resonance Quarterly 9:1–30. 13. Thali, M. J., R. Dirnhofer, R. Becker, W. Oliver, and K. Potter. 2004. Is “virtual histology” the next step after the “virtual autopsy”? Magnetic resonance microscopy in forensic medicine. Magn Reson Imaging 22:1131–38. 14. Lee, R. C. and M. S. Kolodney. 1987. Electrical injury mechanisms: electrical breakdown of cell membranes. Plast Reconstr Surg 80:672–79. 15. Lee, R. C. and M. S. Kolodney. 1987. Electrical injury mechanisms: dynamics of the thermal response. Plast Reconstr Surg 80:663–71. 16. Lee, R. C., D. Zhang, and J. Hannig. 2000. Biophysical injury mechanisms in electrical shock trauma. Annu Rev Biomed Eng 2:477–509. 17. Lee, R. C. 1997. Injury by electrical forces: pathophysiology, manifestations, and therapy. Curr Probl Surg 34:677–764. 18. Rouge, D., A. Polynice, J. L. Grolleau, B. Nicoulet, J. P. Chavoin, and M. Costagliola. 1994. Histologic assessment of low-voltage electrical burns: experimental study with pigskin. J Burn Care Rehabil 15:328–34. 19. Torre, C., L. Varetto, and G. Mattutino. 1986. Dermal surface morphology in wound healing. An experimental scanning electron microscope study. Am J Forensic Med Pathol 7:337–43. 20. Nettelblad, H., K. A. Thuomas, and F. Sjoberg. 1996. Magnetic resonance imaging: a new diagnostic aid in the care of high-voltage electrical burns. Burns 22:117–9. 21. Karczmar, G. S., L. P. River, J. River, et al. 1994. Prospects for assessment of the effects of electrical injury by magnetic resonance. Ann N Y Acad Sci 720:176–80. 22. Fleckenstein, J. L., D. P. Chason, F. J. Bonte, et al. 1995. High-voltage electric injury: assessment of muscle viability with MR imaging and Tc-99m pyrophosphate scintigraphy. Radiology 195:205–10. 23. Ohashi, M., J. Koizumi, Y. Hosoda, Y. Fujishiro, A. Tuyuki, and K. Kikuchi. 1998. Correlation between magnetic resonance imaging and histopathology of an amputated forearm after an electrical injury. Burns 24:362–68. 24. Bittoun, J., H. Saint-Jalmes, B. G. Querleux, et al. 1990. In vivo high-resolution MR imaging of the skin in a wholebody system at 1.5 T. Radiology 176:457–60. 25. Weis, J., A. Ericsson, and A. Hemmingsson. 1999. Chemical shift artifact-free microscopy: spectroscopic microimaging of the human skin. Magn Reson Med 41:904–8. 26. Richard, S., B. Querleux, J. Bittoun, et al. 1993. Characterization of the skin in vivo by high resolution magnetic resonance imaging: water behavior and age-related effects. J Invest Dermatol 100:705–9. 27. Idy-Peretti, I., J. Bittoun, F. A. Alliot, S. B. Richard, B. G. Querleux, and R. V. Cluzan. 1998. Lymphedematous skin
© 2009 by Taylor & Francis Group, LLC
28.
29.
30.
31.
32. 33.
34.
35. 36.
37.
38.
39.
40. 41. 42.
43.
44.
45.
46. 47. 48.
and subcutis: in vivo high resolution magnetic resonance imaging evaluation. J Invest Dermatol 110:782–87. Querleux, B., S. Richard, J. Bittoun, et al. 1994. In vivo hydration profile in skin layers by high-resolution magnetic resonance imaging. Skin Pharmacol 7:210–16. Hinshaw, J. R. and H. F. Pearse. 1956. Histological techniques for the differential staining of burned and normal tissue. Surg Gynecol Obstet 103:726–30. Caffey, J. 1974. The whiplash shaken infant syndrome: manual shaking by the extremities with whiplash-induced intracranial and intraocular bleedings, linked with residual permanent brain damage and mental retardation. Pediatrics 54:396–403. Duhaime, A. C., C. W. Christian, L. B. Rorke, and R. A. Zimmerman. 1998. Nonaccidental head injury in infants— the “shaken-baby syndrome.” N Engl J Med 338:1822–29. Uscinski, R. H. 2006. Shaken baby syndrome: an odyssey. Neurol Med Chir (Tokyo) 46:57–61. Wolfson, D. R., D. S. McNally, M. J. Clifford, and M. Vloeberghs. 2005. Rigid-body modelling of shaken baby syndrome. Proc Inst Mech Eng [H] 219:63–70. Kivlin, J. D., K. B. Simons, S. Lazoritz, and M .S. Ruttum. 2000. Shaken baby syndrome. Ophthalmology 107:1246–54. Kanter, R. K. 1986. Retinal hemorrhage after cardiopulmonary resuscitation or child abuse. J Pediatr 108:430–32. Goetting, M. G. and B. Sowa. 1990. Retinal hemorrhage after cardiopulmonary resuscitation in children: an etiologic reevaluation. Pediatrics 85:585–88. Odom, A., E. Christ, N. Kerr, et al. 1997. Prevalence of retinal hemorrhages in pediatric patients after in-hospital cardiopulmonary resuscitation: a prospective study. Pediatrics 99:E3. Johnson, D.L., D. Braun, and D. Friendly. 1993. Accidental head trauma and retinal hemorrhage. Neurosurgery 33:231–34, discussion 234–35. Betz, P., K. Puschel, E. Miltner, E. Lignitz, and W. Eisenmenger. 1996. Morphometrical analysis of retinal hemorrhages in the shaken baby syndrome. Forensic Sci Int 78:71–80. Bradley, W. G., Jr. 1993. MR appearance of hemorrhage in the brain. Radiology 189:15–26. Roob, G. and F. Fazekas. 2000. Magnetic resonance imaging of cerebral microbleeds. Curr Opin Neurol 13:69–73. Blamire, A. M., J. G. Rowe, P. Styles, and B. McDonald. 1999. Optimising imaging parameters for post mortem MR imaging of the human brain. Acta Radiol 40:593–97. Petiet, A., L. Hedlund, and G. A. Johnson. 2007. Staining methods for magnetic resonance microscopy of the rat fetus. J Magn Reson Imaging 25:1192–98. Spencer, R. G., K. W. Fishbein, A. Cheng, and M. P. Mattson. 2006. Compatibility of Gd-DTPA perfusion and histologic studies of the brain. Magn Reson Imaging 24:27–31. Sodickson, D. K. and C. A. McKenzie. 2001. A generalized approach to parallel magnetic resonance imaging. Med Phys 28:1629–43. Larkman, D. J. and R. G. Nunes. 2007. Parallel magnetic resonance imaging. Phys Med Biol 52:R15–5. Katscher, U. and P. Bornert. 2006. Parallel RF transmission in MRI. NMR Biomed 19:393–400. Setsompop, K., L. L. Wald, V. Alagappan, et al. 2006. Parallel RF transmission with eight channels at 3 Tesla. Magn Reson Med 56:1163–71.
Internal Body Documentation
93
B2.4.3 NUCLEAR MAGNETIC RESONANCE SPECTROSCOPY IN FORENSIC MEDICINE Eva Scheurer, Peter Bigler, Michael Ith, and Chris Boesch B2.4.3.1 Introduction: Nomenclature: In Vitro versus In Vivo Applications of NMR Spectroscopy Nuclear magnetic resonance (NMR) is a physical effect that can be used in very different ways, leading from morphological images|as shown in most chapters of this book|to magnetic resonance spectra in situ and in vitro. The basics of this effect are explained in Section B2.4.1, Basics of MR Imaging and Spectroscopy. Additionally, several textbooks and overviews that give an introduction into the basics of magnetic resonance spectroscopy can be recommended [1–6]. While the same physical effect is used, it is nevertheless important for practical reasons to distinguish in situ and in vitro spectroscopy. In the following, in situ spectroscopy will be called “magnetic resonance spectroscopy” (MRS), while in vitro spectroscopy methods are either called “high-resolution NMR” in solutions or “magic angle spinning” (MAS) in tissues. The major difference is that MRS is performed noninvasively on the animal or human body, while high-resolution NMR investigates pure solutions or mixtures in test tubes. As a consequence, the available magnet field strengths of the two methods and therefore also sensitivity and resolution of their spectra are markedly different. Unlike magnetic resonance imaging (MRI) that provides morphological images predominantly based on the proton (1H) signal of water, MRS and high-resolution NMR observe multiple compounds and metabolites using the signals from various isotopes such as hydrogen (1H), phosphorus (31P), carbon (13C), fluorine (19F), sodium (23Na) and others. Table B2.4.3.1 summarizes typical differences between the various spectroscopy methods.
B2.4.3.2 In Vitro NMR Spectroscopy: High-Resolution NMR and MAS Using in vitro NMR spectroscopy, small samples of liquids can be studied nondestructively, leading to a wealth of information about the sample. As compared to optical methods, NMR suffers from an inherently low sensitivity, which makes the method inappropriate for molecular trace analysis. On the other hand, NMR offers great advantages for structure elucidation of pure samples or−most promising for forensic medicine−for the analysis of mixtures of compounds. Today, high-resolution NMR is an indispensable method for chemical labs, for pharmaceutical companies, and for structure analysis in biophysics, based on a comprehensive arsenal of NMR experiments. The magnets used for high-resolution NMR have a vertical bore, adequate for probeheads (i.e., radio frequency coils and additional hardware) dedicated for sample tube diameters of up to 10 mm. Spectrometers with magnetic field strength of about 4.7 to 21.14 Tesla, corresponding to a proton frequency of 200 to 900 MHz, are available. An NMR sample tube is easily prepared by dissolving the sample (typically a few mg) in an adequate deuterated solvent, or for aqueous samples (typically 400 μl) by adding a small volume (about 200 μl) of deuterium oxide (D2O). Deuterated solvents are used to keep the proton signal of the solvent low and serve for spectrometer stabilization and optimization of the field homogeneity. Since the solvent can interact with the NMR signals, which is often the case in 1H-NMR spectra, the choice of the solvent is crucial. With the addition of a known amount of a standard, i.e., tetramethylsilane (TMS) or sodium trimethylsilylpropionate (TSP), the components of a sample can be quantified in relative or absolute unities. Usually, a few milligrams of a sample are used for an NMR investigation on a routine level; however, sample amounts below 1 μg are also feasible taking advantage of newest sensitivity driven technologies.
TABLE B2.4.3.1 Characteristic Values for the Application of High Resolution NMR Spectroscopy, Magic Angle Spinning (MAS), and In Situ MRS In Vitro NMR and MAS typical field strength (in Tesla) typical magnet bore size typical proton resonance frequency sample volume
up to 21.14 T few centimeters 200–900 MHz about 0.5 ml
sensitivity
typically 1-20 mg/500 Ml, a few Mg/500 Ml with special equipment min–hours liquids (high-resolution NMR), tissue (MAS)
acquisition time observed material
© 2009 by Taylor & Francis Group, LLC
In Situ MRS 1–3 T up to 1 meter (42–128 MHz) 1–4 ml (1H), up to 100 ml for insensitive isotopes such as 13C 1 mM min–hours tissue
94
In contrast to other analytical methods, NMR is best suited for the investigation of mixtures of compounds since there is no need for a compound separation, e.g., by chromatography, prior to the NMR investigation. However, solid structures such as tissues or cells have to be removed or extracted first; otherwise the magnetic field homogeneity is affected and no high resolution spectra may be obtained. There is one technique called magic angle spinning (MAS), which overcomes this problem with very fast spinning of the sample in the magnetic field at a specific angle, requiring special equipment. However, without MAS an extraction of the solid parts in the sample is necessary. Depending on the solubility of the various compounds of interest, different solvents are proposed for aqueous and organic metabolite fractions. For the water-soluble part, perchloric acid (PCA) extraction is commonly used [7–10], while the methanol-chloroform-water (M/C) technique is well established for lipid extractions of cells and tissues [11]. Particularly interesting are methods for the simultaneous extraction of lipids and water-soluble metabolites [12–14]. Since NMR is nondestructive, a repetitive analysis of the same sample is possible, e.g., to investigate the progress of metabolism. The acquisition time depends on the amount of the investigated sample and its concentration in the NMR tube, but also on the sensitivity of the measured isotope and the complexity of the chosen NMR experiment. It would go far beyond the scope of this chapter to summarize sophisticated NMR experiments (e.g., 2D COSY, 2D HSQC, and others); however, it should be mentioned that these methods provide an enormous amount of information about the composition of mixtures and the structure of molecules. Spectral processing and interpretation is strongly dependent on the type of experiment. Generally, a one-dimensional proton spectrum can be acquired and processed in a few minutes, while multi-dimensional spectra take much more time for both. B2.4.3.3 In Situ MRS In contrast to in vitro NMR spectroscopy, in situ 1H-magnetic resonance spectra (1H-MRS) can be obtained on standard clinical MR systems with typical field strength of 1.5 to 3 Tesla. Thus it is possible to examine the chemical composition of selected volumes in humans non-invasively [1,3,15,16]. The MRS signal acquisition is restricted to a well-defined volume (“voxel,” “region of interest ROI”), which can be placed at the desired anatomical location based on a series of scout MR images of the region. Alternatively, the so-called chemical shift imaging (CSI) where a virtual grid of voxels is placed onto an organ or a tissue allows the locally resolved analysis of the chemical composition of an entire anatomical region. Particularly interesting for forensic medicine is the examination of living people with MRS, especially if MRS is combined with MRI. Compared to high-resolution NMR, MRS has much lower sensitivity and the spectral resolution is limited. As there is no radiation exposure, MRS can be applied to measure the metabolic state of certain tissues and
© 2009 by Taylor & Francis Group, LLC
The Virtopsy Approach
organs non-invasively and—if required—even repetitively. In general, a single spectrum can be acquired by technicians in fractions of an hour without the need of sample preparation, and contains information on about 20 metabolites simultaneously. Depending on the state of automated spectral analysis, final data can be obtained briefly after the data acquisition. B2.4.3.4 NMR Spectroscopy and Pertinent Issues in Forensic Medicine In Table B2.4.3.2, some questions are listed that could be of help planning a study or just a single examination with MR spectroscopy. In any case, it is highly recommended to discuss the project already in an initial state with an (N)MR specialist. The assistance of technicians, physicists, or chemists is essential for the realization of the measurements and the analysis of the data. Forensic medicine is concerned with problems occurring in the judicature which can only be solved by scientific methods and knowledge in medicine, biology, chemistry, and physics. The arising questions may concern living or deceased persons. Table B2.4.3.3 gives an overview of the most frequent questions that arise for the forensic pathologist. Printed in bold are the issues where the application of NMR spectroscopy could be useful. Since NMR spectroscopy delivers chemical information, forensic medicine could particularly benefit from information concerning metabolism and toxicology, i.e., by measuring metabolite or substance concentrations or the composition of tissues and fluids. Antemortem metabolism can be investigated in living people, answering questions on injuries and diseases associated with abnormal metabolite concentrations, or it could be used for evaluation of the aging processes. MRS is not associated with any harmful radiation; thus informed consent for the participation of volunteers is generally easily achieved.
TABLE B2.4.3.2 Decision Guidance Whether and in Which Form MR Spectroscopy Could Be Utilized Shall the measurement be done non-invasively and in situ (i.e., MRS) or are there excised material or body fluids (i.e., high-resolution NMR or MAS) available? What kinds of MR systems are available? (e.g., whole body versus analytical NMR systems, type of coils, observable nuclei, etc.) Would there be an advantage by combining chemical and anatomical information, i.e., by combining MR spectroscopy with MR imaging? What collaborations with (N)MR groups are possible? (Need for expert knowledge) What is the aim of the investigation or study? Are the persons to be examined living or deceased? What substances of interest are expected in the sample and which are their supposed concentrations? Shall the examination give information on the normal metabolism, pathological metabolism or on external substances?
Internal Body Documentation
TABLE B2.4.3.3 Issues in Forensic Medicine in General and in Particular Those Issues Which Could Possibly Benefit from NMR Spectroscopy Living Persons:
Deceased Persons:
Sexual assault Child abuse Criminal assault Age determination Abuse of drugs/alcohol Paternity
Identification Manner of death Time of death, postmortem interval Cause of death Injury by different forces Vitality of injuries Reconstruction, causality Toxicology, abuse of drugs/alcohol
Postmortem metabolite concentrations can provide information on pathology that leads to the death of the person or can serve as biomarkers for certain antemortem conditions and causes of death. As with MRS, noninvasive repetitive measurements of a precise volume of tissue are possible postmortem decomposition processes that can be followed without destroying any evidence. Additionally, high-resolution NMR as well as MRS offers a great potential for the investigation of criminalistic questions such as the examination of explosives, hydrocarbon fuels, lachrymators, and particularly mixtures of different compounds [17]. Potential applications of NMR spectroscopy in forensic medicine are evaluated next. The examples, which are in no way complete, do not represent the daily routine, and some of them would probably not even pass a feasibility study. However, these examples are summarized to show the opportunities of NMR in forensic medicine. Since NMR has developed with an incredible speed, many potential applications may be feasible now since sensitivity and spectral resolution have been greatly improved and could add to the repertory of forensic methods. B2.4.3.5 Potential Applications of NMR Spectroscopy to Living Persons MRS in living persons or NMR spectroscopy of body fluids offer many advantages compared to either imaging modalities alone or to conventional chemical methods. Highresolution NMR spectroscopy can be applied to any kind of homogenous body fluid, e.g., serum, CSF, or urine, without any sample preparation apart from adding a small volume of D2O. Thus the sample remains unchanged by the NMR examination and can be re-examined with other methods, or can be stored as evidence. The measurement of blood or tissue samples must in general be preceded by an adequate extraction. As explained above, NMR may be performed with very small sample amounts, which is particularly advantageous with children where sample volumes are often restricted. Though the sensitivity is moderate compared to the detection
© 2009 by Taylor & Francis Group, LLC
95
limits of chromatographic methods, e.g., GC-MS, NMR is particularly useful to get a survey of all compounds contained in the sample and for their accurate and simple quantitation. Depending on the question, different tissues, e.g., brain, liver, muscle tissue, kidney, bone, or adipose tissue can be investigated by in situ MRS. Applied together with MR imaging in a single examination, metabolic or chemical information can be combined with precise anatomical localization. This allows study of the content in toxic substances, drugs of particular pathology, or injured regions of the body, as well as investigation of characteristic metabolic processes in situ. As the sensitivity is relatively low, attention must be paid to the expected concentrations of the substances of interest. Depending on the examined metabolites and substances in situ, MRS as and NMR can both be applied to different nuclei, e.g., 1H, 13C, 19F, 31P. B2.4.3.5.1
Child Abuse
Child abuse, and particularly abusive head trauma, is an important issue in forensic medicine as it concerns almost a quarter of all children less than 3 years of age who are admitted to hospitals [18]. With an incidence of serious or fatal inflicted traumatic brain injury (iTBI) of approximately 1 in 3300, children of less than 1 year are most affected [19]. However, even with a severe brain injury children at that age often present with nonspecific symptoms, e.g., vomiting without diarrhea, poor feeding, crying, or fussiness, which makes a rapid diagnosis difficult. The distinction of children with benign causes for their symptoms from such with traumatic brain injury is challenging, particularly in inflicted TBI where a clear and adequate history and a known time of injury is mostly missing [20–23]. Misdiagnosis or delayed diagnosis of iTBI is common and results in increased morbidity and mortality [23,24]. Biochemical markers of injury are routinely used to assist in the diagnosis of organ injury, e.g., troponin in myocardial infarction. For the detection of brain injury a number of molecules have been proposed as biomarkers as they are released from brain tissue after injury. It has been suggested that the molecules pass into the cerebrospinal fluid (CSF) and to a certain extent cross the blood−brain barrier to be detected in serum, and finally in urine [25]. A number of studies address biomarkers to facilitate the diagnosis of iTBI. In CSF the concentration of cytochrome c, a biomarker of apoptosis, was compared in infants and children with TBI, including inflicted trauma cases to controls without trauma or meningitis [26]. Increased CSF cytochrome c was independently associated with iTBI and female gender, but not age, GCS, or survival. This result suggests that apoptosis plays an important role in the subset of TBI patients diagnosed with child abuse. Quinolinic acid is a metabolite of tryptophan metabolism produced by macrophages and microglia associated with inflammatory response in the central nervous system (CNS). A powerful association between CSF quinolinic acid and time after injury could be shown for adult patients with severe TBI
96
(Glasgow Coma Scale [GCS] < 8) [27]. After the first 12 h-period with normal levels of quinolinic acid, a steady increase up to a maximal concentration after 72 to 83 h after injury was reported. Additionally, after correcting for the effect of time an association was found with prognosis, i.e., patients who died had higher levels of quinolinic acid versus survivors, while no correlation was found with age, gender or initial GCS. In children (2 m−16 y) the significant association of quinolinic acid concentration in CSF with time after TBI was confirmed [28]. Peak concentrations were found after 48–72 h after injury for both inflicted and accidental TBI [22]. In a comparison of iTBI cases with controls with accidental traumatic brain injury, the patients with iTBI had significantly increased initial and peak CSF concentrations of quinolinic acid [22].2 As quinolinic acid concentration seems to be an indicator of time since injury, this observation would suggest that SBS cases are presented to the hospital with a significant delay, possibly resulting in enhanced secondary injury. Additionally, the increased concentrations could be related to the more severe injury and the younger age of the concerned children [22]. Although it was suggested that markers in CSF would have a higher sensitivity and specificity for detecting iTBI than in serum because of the anatomical closeness to the brain, the difficulty of obtaining CSF from infants in an outpatient setting makes it rather unlikely to be useful as a screening test. In serum, the use of biomarkers to detect brain injury has been extensively studied. For a pediatric population neuronspecific enolase (NSE), S100B, and myelin-basic protein (MBP) have been proposed. As NSE, a glycolytic enzyme localized primarily to the neuronal cytoplasm and used as a neuronal marker, is also present in erythrocytes and platelets, the sample needs to be non-hemolyzed. The initial serum concentration of NSE can detect traumatic brain injury with a sensitivity of 71% and a specificity of 64%, while S100B, a calcium-binding protein localized to astroglia, is reported to be more indicative of intracranial injury than findings on a CT scan with 77% sensitivity and 72% specificity [29]. A combination of the initial concentrations of NSE and S100B even led to 80% sensitivity and 73% specificity for the identification of TBI. However, in this study the only differentiation of nTBI from iTBI lies in a significant difference in the time to peak concentration, whereas the concentrations of the biomarkers themselves is not different. A comparison of the time course of NSE, S100B, and MBP after hypoxic-ischemic brain injury, iTBI and nTBI suggests that the biochemical response in iTBI is distinct from nTBI with temporal similarities with hypoxic-ischemic injury [30]. In a prospective screening of children <1 year presenting with nonspecific symptoms and no trauma history, NSE was 77% sensitive and 66% specific, and MBP was 36% sensitive and 100% specific for iTBI [24]. An increase in serum MBP is reported to be specific for intracranial hemorrhage and severe TBI. The concentration peaks approximately 72 h after injury and can remain increased for up to 2 weeks [31]. This, in contrast to other markers’ late peak, may assist in determination of the timing of injury. For the prediction of intracranial
© 2009 by Taylor & Francis Group, LLC
The Virtopsy Approach
hemorrhage, a sensitivity of 44% and a specificity of 96% was assigned [29]. S100B has also been measured in urine, where it peaks significantly later than in serum [25].2 In 80% of the subjects with acute brain injury (iTBI and hypoxemic brain injury) increased serum and urinary S100B could be detected, but in none of the control patients. The detection of urinary S100B also correlated with a poor outcome (Glasgow Outcome Scale between 6 and 12 months after injury) with a sensitivity of 83% and a specificity of 100%. In conclusion, the first and most important goal, not only in forensic medicine but also for the pediatric clinic, is to establish a battery of several biomarkers as an assistant tool in the detection of traumatic brain injury. A second step involves the screening of children <1 year, who are at highest risk for suffering from inflicted traumatic brain injury and additionally, where diagnosis is challenging. However, though a number of studies have already been performed on investigating this subject, more data is needed, particularly on the specificity concerning the differentiation against nontraumatic neurologic diseases, e.g., seizures, meningitis, or hypoxic-ischemic encephalopathy, and on sensitivity of the markers when using a more sensitive gold standard to intracranial injury than CT, e.g., MRI [32–35]. The measurements of the biomarkers in all the studies so far have been made by ELISA (enzyme-linked immunosorbent assay) which is elaborate and time-consuming. By applying high resolution NMR, several markers could be detected simultaneously in a single spectrum, saving time and costs. Additionally, markerspecific spectral patterns in the corresponding spectra can be evaluated and may be used to simplify their recognition in subsequent investigations. The rapid diagnosis of inflicted brain injury is of great importance for adequate medical treatment as well as prognosis and outcome, which in turn is of particular forensic importance. The so-called Shaken Baby Syndrome (SBS) is the most common concept of death or serious neurological injury resulting from child abuse characterized by subdural and retinal hemorrhages, sometimes accompanied by occult bone fractures [36]. However, the diagnosis is usually difficult, since none of these findings is pathognomonic of SBS or needs to be present at all [34]. To date, it is predominantly based on a constellation of clinical findings in the absence of an adequate history, frequently completed by a computed tomography of the head and an ophthalmologic examination. However, MRI seems to be more sensitive for the detection of many intracranial findings in infants than CT (e.g., subdural hemorrhages) and offers the advantage that blood accumulation can be more accurately dated, thus helping to determine the timing of the injury [35,37–41]. In addition, diffusionweighted MRI has proven to be useful in the diagnosis of acute cerebral ischemia within minutes of onset, and the detection of diffuse axonal injury [42–44]. The combination of MRI with MRS offers the possibility to make prognostic statements on long-term neurologic outcome. It seems that at the early time of about 5 days after injury, proton spectra show clearly different patterns of abnormality, particularly
Internal Body Documentation
97
A
NAA Cho Cr mI Lac/Lip
B
C
4
3
ppm
2
1
0
FIGURE B2.4.3.1 In vivo proton MR spectra from the parieto-occipital cortex of infants A (at 7 days posttrauma), B (5 days posttrauma), and C (7 days posttrauma). While the spectrum from infant A appears normal, the spectra from B and C show major abnormalities. B: Markedly reduced NAA, low Cr, appearance of signals in the Lac/Lip regions; C: almost absent NAA, low Cr and mI, and dominant peaks in the Lac/Lip regions [45].
in the levels of N-acetyl aspartate (a neuronal marker), creatine and phosphocreatine, lactate and lipids [45]. In Figure B2.4.3.1, in vivo proton spectra of three infants (A: 6 months, B: 5 weeks, C: 7 months) with clinically established diagnosis of SBS are shown. While up to 19 days posttrauma little or no changes indicating the severity of sustained neuronal injury were detected by MR imaging, spectra taken as early as 5 to 7 days posttrauma provide a guide as a prognostic indicator for long-term outcome. Infant A recovered completely, while B and C remained in a neurologically severely compromised state. The pattern of metabolic changes shows some similarities with hypoxic injury, but also seems to have different
© 2009 by Taylor & Francis Group, LLC
com-ponents [46,47]. Using peak area metabolite ratios and the presence of lactate with a long echo time protocol (TE = 270 ms), correct outcome predictions for 6 months after injury were possible in 90% to 100% of the children (3 d−18 y) suffering from acute brain injury [48]. Even lactate alone being correlated with hypoxic-ischemic encephalopathy was able to predict the outcome at 6 to 12 months after discharge in 96% of infants and children (1 m−15 y) with severe closed head injury from non-accidental or accidental incidents [49]. The same observation was the result of a small prospective study of 11 infants with inflicted TBI [50]. Larger populations need to be studied, focusing on the early detection of inflicted head injury and the identification of markers for early prediction of
98
clinical outcomes. This might help not only to guide medical treatment and rehabilitative management of these patients, but also for early initiation of social and forensic measures.
B2.4.3.6 The Application of MRS to the Problem of Age Determination In recent years the assessment of age in living subjects has become an important part in forensic practice, particularly as a result of cross-border migration. Depending on the cause why the genuine age needs to be clarified, i.e., in the course of criminal, civil, asylum, or old-age pension proceedings, the relevant age thresholds lie between 14 and 22 years and at 60 or 65 years, respectively. To date in Europe, the state of the art has been proposed by the interdisciplinary Study Group on Forensic Age Diagnostics as a combination of a physical examination with recording of anthropometric data, signs of sexual maturation, and age-relevant developmental disorders, an x-ray examination of the left hand, and a dental examination with recording of the dentition status and the evaluation of an orthopantomogram (OPG) [51–54]. For the assessment of the age threshold of 18 years and over, an additional radiographic or CT examination of the clavicles is recommended [55]. The reliability of the individual methods and certain combinations was evaluated in various studies [56–58]. They conclude that the results of the physical examination do not allow age estimation, but are only useful to rule out the existence of pathological conditions that may alter maturity (nocturnal enuresis, GH deficit, practice of competitive sports, skeletal malformations, effects of physical agents, e.g., cold). The anthropometric figures and classification in stages, e.g., the Tanner stages of secondary sexual characteristics, are very imprecise and can only be used for extreme statements, i.e.,” is definitely older than...” Though the skeletal age obtained from x-ray of the hand is not absolutely correlated to the subjectsí actual chronological age, it is nevertheless the main parameter for age estimation at the end of adolescence with some accuracy, being considered physiologically more stable than dental maturation [56]. A comparison of different dental age estimation methods using radiographs of developing teeth of living children between 3 and 17 years showed that the most accurate method for both boys and girls was the method by Willems, with mean differences between chronological age and dental age of −0.05 (boys) to −0.2 years (girls), i.e., resulting in a slight underestimation [58,59]. An evaluation of orthopantomograms based on a combination of different features concerning the question of whether a person has attained 21 years of age resulted in a probability of correct classification of 69.7% for males and 71.4% for females 18−30 years of age [57]. According to the authors, this would not be sufficiently accurate for criminal proceedings, thus they recommend the additional radiographic examination of the clavicle. A combination of an evaluation of skeletal age with the assessment of dental maturity, i.e., of the two x-ray based methods, resulted in a significant increase in the efficacy of
© 2009 by Taylor & Francis Group, LLC
The Virtopsy Approach
the prediction of whether the 18-year-old age limit had been exceeded or not. Thus the accuracy of the proposed procedure with external examination, radiography of the hand, and OPG stands mainly on the x-ray-based methods. Depending on the country-specific legislation, the application of x-raybased examinations is problematic, particularly for noncriminal issues such as civil or asylum proceedings. In old-age pension cases the chronological age of mature persons of 40 years and over has to be estimated where the skeletal and dental development has long been terminated. For these cases, the interdisciplinary Study Group on Forensic Age Diagnostics proposes the evaluation of radiological data of medical examinations performed in childhood or adolescence. Unfortunately, appropriate radiographies will only be available in rare cases. A biochemical age estimation by the means of the degree of racemization of asparagic acid in dentin is far more exact, but needs a dental extraction, and thus is connected with ethical problems and legal questions. A quantitative method not based on the maturity of certain tissues and free of radiation exposure would well serve the purpose of forensic age estimations in living persons of different chronological age. MR-based methods are free of radiation exposure and can be used for examinations of persons of any age, as far as
A
A
(A)
FIGURE B2.4.3.2 (A) Sagittal T2-weighted MR image of lumbar spine with voxel position.
Internal Body Documentation
99
0.5
Water
0.3
0.2
Intensity (a.u.)
0.4
Fat 0.1
0.0
200
100
0
–100
–200
–300
–400
–500
Frequency (Hz) (B) Relative Fat Intensity in Red Vertebral Bone Marrow 70%
Fat (% total intensity)
60% 50% 40% 30% 20%
Male
Female
10% 10
20
30
40
50 60 Age (years)
70
80
90
100
(C)
FIGURE B2.4.3.2 (B) In vivo proton spectrum from a 34-year old female volunteer with relative fat intensity 23.9% (corrected for an echo time (TE) of 0 ms). (C) Relative fat intensity in the red lumbar bone marrow versus age in male (x) and female (o) volunteers. Note that forensically relevant age groups are not well represented [60].
some security aspects are considered (e.g., implanted medical electronic and magnetic metal devices). Based on a clinical approach, several studies found a linear correlation between age and fat content of the hematopoietic bone marrow of the lumbar vertebrae based on a gradual, steady, and progressive change from hematopoietic to fatty marrow [60–63]. Using in situ 1H-MR spectroscopy, the fat content was measured in a relatively short examination, i.e., about 30 minutes in total. In Figure B2.4.3.2A, voxel positioning in the lumbar spine is shown. As the fat and water signal are present in the same spectrum (see Figure B2.4.3.2B), the fat fraction or the fat−water ratio can be calculated easily. Generally, the lipid fraction is found to increase linearly about 5−7% per decade of age, with a low fat fraction of 15−24% in the age group
© 2009 by Taylor & Francis Group, LLC
from 11−20 years and a high fat fraction of 54−55% in persons 61−70 years. As an example, the results for the measurement of the relative fat intensity in red vertebral bone marrow of the study of Kugel et al. are shown in Figure B2.4.3.2C [60]. However, because of the different aims of the clinical studies, the forensically most interesting age groups, i.e., between 14 and 22 years for the adolescent thresholds and between 55 and 70 years for old-age-pension problems, have not been sufficiently included. Thus the results need to be verified in studies examining a greater number of healthy volunteers in the concerned age groups. An important factor influencing the fat fraction of the bone marrow is gender. Most studies reveal that females have a significantly reduced fat fraction
100
in comparison to males. Though this difference has been reported in all age groups, it is most pronounced in the 4th and 5th decade, while it is reduced to only 1% for the age group of 11−20 years and for persons over 60 years [60,62]. There was found to be no correlation of relative fat content with body mass index. However, the relation of the fat fraction of the bone marrow with diseases of the hematopoietic system as well as diseases correlated with a change of bone density is evident and could easily be confirmed [62,64]. B2.4.3.7 Abuse of Drugs or Alcohol B2.4.3.7.1 High Resolution NMR Spectroscopy of Biofluids Used for the Diagnosis of Acute Poisoning The application of high resolution NMR spectroscopy to forensic analytical toxicology has been proposed by many authors, starting in the early days of NMR applications in 1966 [65]. Compared to the routinely used techniques requiring extraction hydrolysis of conjugated forms, chromatographic methods, and sometimes derivatization, the main advantages of NMR are the simple sample preparation and the small sample volumes (300−500 μl). A wide range of xenobiotics can be characterized and quantitated directly in urine and serum. Examples of acute poisoning cases where urine and serum samples have been analyzed with 1H- and 13C-NMR include salicylate, valproic acid (therapeutic drugs), paraquat (a herbicide), tetrahydrofuran (a solvent), methanol, and ethylene glycol [66–70]. Two-dimensional NMR techniques can be applied to disentangle regions in one-dimensional spectra with large signal overlap, and can be used to visualize spin−spin coupling interactions. This allows spectral analysis to be simplified for easier identification of substances and for obtaining additional information on metabolites and biotransformation of xenobiotics. For example, J-resolved 1H NMR spectroscopy, dedicated to separate chemical shift and coupling information in different dimensions, leads to spectral simplification and characterization of low molecular weight compounds. This technique has been used in acute intoxication cases on urine samples in addition to the onedimensional experiment to analyze the metabolites of acetylsalicylic acid and chloroquine, an antimalaria drug [71–74]. Figure B2.4.3.3 shows, in addition to the molecular structure of chloroquine, an example of the application of NMR spectroscopy to a case of attempted suicide of a 41-year-old man. The one-dimensional 1H spectrum of the crude urine shows intense peaks between 1.2 and 1.5 ppm and near 8.3 ppm, which are not usually present in the urine of a normal healthy subject. Two expansions of the two-dimensional J-resolved spectrum allow for an assignment of the unusual signals to chloroquine protons. Additionally, signals for major endogenous metabolites (e.g., lactate, alanine, 3-D-hydroxybutyrate, hippurate) as well as ethanol could be identified in the urine spectra. Other two-dimensional sequences used in acute toxicological screenings include 1H-1H-TOCSY (Total Correlation
© 2009 by Taylor & Francis Group, LLC
The Virtopsy Approach
Spectroscopy) and 1H-13C-HMBC (Heteronuclear Multiple Bond Coherence) [67,73]. A combination of 1H and 31P NMR spectroscopy is used for the detection and quantitation of glyphosate, an organophosphorus compound used worldwide as a herbicide [75]. 31P NMR is particularly useful in this case, as endogenous phosphorus metabolites present in plasma or urine show no signal in the considered chemical shift region. To avoid interpretation problems by overlapping signals of therapeutic agents, collection of quantitative information from spectra recorded from biological samples taken before any treatment is recommended. In clinical emergency situations, NMR spectroscopy is considered to be useful for rapidly confirming the diagnosis of poisoning and in evaluating the effectiveness of therapeutic elimination procedures. 1H NMR spectroscopy has also been successfully applied to observe legal supplementation as well as illegal doping practices. The kinetics of creatine supplementation could be characterized by measuring the evolution of urinary creatine excretion after oral intake [76]. The investigation of possibly adulterated ingested beverages is often problematic because there is no previous knowledge what substances to search for. 1H NMR spectroscopy has proven to be useful in a case of intentional poisoning by adulterated soda where several added substances as well as endogenous metabolites highlighting metabolic disturbances were found [77]. In another case, serum and urine of a comatose woman whose husband was the owner of a dry cleaning shop was investigated by 1H NMR. The spectra (see Figure B2.4.3.4) show a characteristic signal at 1.9 ppm, which could be attributed to protons of THF (two D-methylene groups). Additionally, unusual peaks not appearing in control samples led to the unexpected detection of GHB (I-hydroxybutyrate) [71]. B2.4.3.7.2 Detection and Quantification of GHB Gamma-hydroxybutyrate (GHB) is a powerful illicit drug acting on the central nervous system. While at low doses GHB can cause a state of euphoria, increased libido and sociability, at higher doses depressant effects including nausea, dizziness, drowsiness, agitation, visual disturbances, depressed breathing, amnesia, and unconsciousness prevail. The effects of GHB can last from 1.5 to 3 hours, or even longer if large doses have been consumed or if it is mixed with alcohol. At high concentrations GHB acts as a sedative and hypnotic with a very narrow therapeutic index, thus presenting a high risk for overdose, and has an amnesic effect. Recreationally used GHB comes in the form of a colorless, odorless liquid or as a powder-like sodium salt with a thin, very salty, chemical taste. The use of GHB has gained popularity in the dance/rave scene and has been reported in drug-facilitated sexual assault, where it is added to beverages. GHB is rapidly eliminated from the body, i.e., the half-life in blood plasma is reported to be 30 to 60 minutes after low to moderate single oral doses [78–80], and less than 5% of the ingested GHB is excreted unchanged in the urine. Thus recognition and detection of GHB in cases of criminal usage is very difficult. Additionally, the fact that
Internal Body Documentation
101
12
A H
CH3
14
CH
CH2
N 11
CH2
5
CH2 15
13
6
CH2
CH3
N CH2
CH3
16
17
Chloroquine
3
C??
N
8
2
C
HOD G Hip
T Ci
U
A * *
* 8
7
6
5
4
3
2
1
(ppm) B Al
H12
H173-HB Eth
L
–10
0
(Hz)
–5
5 10 1.45
1.40
1.35
1.30 1.25 (ppm)
1.20
1.15
1.10
Hip
H2 H5
H3
C
–20
0
(Hz)
–10
10 20 8.4
8.2
8.0
7.8
7.6 (ppm)
7.4
7.2
7.0
6.8
FIGURE B2.4.3.3 A: Structure of chloroquine and 300 MHz one-dimensional 1H NMR spectrum of urine (pH 7.0, NH4Cl 0.8 mol/l); B: Expansion of the aliphatic region (0.7−1.5 ppm) of the 2D J-resolved spectrum with chemical shifts (horizontal axis) and homonuclear protonproton couplings (vertical axis) separated; C: Expansion of the aromatic region (6.7−8.5 ppm) of the J-resolved spectrum. Assignments of the resonances: A acetate, Al alanine, C creatinine, Ci citrate, Eth ethanol, G glycine, 3-HB 3-hydroxybutyrate, Hip hippurate, HOD residual water peak, L lactate, T trimethylamine-N-oxide, U urea, * unusual resonances, Hx chloroquine protons.74
GHB is naturally present in the body complicates the differentiation between endogenous and ingested GHB [81,82]. The standard method for analysis of GHB levels in urine is gas chromatography with flame ionization or mass spectrometric detection, which needs multiple preceding extraction steps
© 2009 by Taylor & Francis Group, LLC
and a sample volume of 2 ml of urine [78]. 1H NMR spectroscopy has been successfully applied to analyze human saliva as well as a low-alcohol beer in order to detect and quantify GHB unequivocally [83]. A sample volume of 0.6 ml was used; sample preparation consisted simply of adding a small
102
The Virtopsy Approach
A CH3 Lac
CH2α GHB
CH2β THF CH2β GHB
CH3 Creat CH2 Ci
TSP-d4
CH3 Ala
3.0
2.5
2.0
1.5
1.0
0.5
0.0
CH2β THF
B
CH3 Lac
CH2α GHB
3.0
2.5
CH2β GHB
2.0
1.5 ppm
TSP-d4
1.0
0.5
0.0
FIGURE B2.4.3.4 1H NMR spectrum of urine (A) and serum (B) from a patient intoxicated with tetrahydrofuran (THF). By chance, I-hydroxybutyrate (GHB) was found in both samples in addition [71]. Signal assignments: THF tetrahydrofuran, GHB 4-G-hydroxybutyric acid, Creat creatinine, Ci citrate, Ala alanine, Lac lactate.
volume of deuterium oxide (D2O), providing a field frequency lock, and TSP (trimethylsilyl propionate) as a chemical shift reference. As shown in Figure B2.4.3.5, the detection of GHB in the 1H NMR spectra of saliva was based on three resonances, i.e., protons ascribable to the C-, D- and I-CH2 groups located at 2.25, 1.81, and 3.61 ppm, respectively, while the I-CH2 proton resonance overlapped with those of carbohydrates in beer and thus was not visible. The authors conclude that high-resolution NMR spectroscopy qualifies well for the detection and quantification of many xenobiotics in miscellaneous biofluids offering many advantages over alternative analytical techniques currently employed, e.g., no invasive sample preparation, ability to simultaneously screen complex, multicomponent matrices, detection of drug precursors and intermediates, and additional diagnostic information regarding the source of illicit drugs from the remaining impurities.
© 2009 by Taylor & Francis Group, LLC
B2.4.3.7.3 Alcohol and its Metabolites The detection and quantification of ethanol and its metabolites in urine by 1H NMR spectroscopy has been well documented in literature [84]. Although in clinical and forensic toxicology the use of urinary ethanol concentration is less important compared to blood and breath analysis, a critical review concludes that urine is a practical and useful alternative if certain precautions are taken, i.e., avoiding manipulation of the specimen or including a chemical preservative in the sampling tubes [85]. A very small amount of the consumed ethanol (less than 1%) is excreted in the urine as ethyl glucuronide, a non-oxidative water-soluble metabolite of ethanol formed in the liver. Owing to the markedly prolonged elimination time compared to ethanol itself, the measurement of ethyl glucuronide in urine can be used as a sensitive and specific biomarker for alcohol intake and for estimation of the time
Internal Body Documentation
103
A
A
Lac-CH3 TAG-(CH2)2– Prop-CH3 NAG-NHCOCH2 Prop-CH2 Pyr Suc Tyr??
TAG-CH3
Form
Lac-CH
Tyr Phe
9.0
8.5
8.0
7.5
??
7.0
6.5
6.0
5.5
5.0
4.5
4.0 ppm
Lys-??-CH2 Gly-CH2 Chol TMAO
Ala-CH3
TSP
3.5 3.0 2.5 2.0 1.0 1.5 PUFA CH-CH-CH2-CH-CH Bμ-β-CH2
0.5
0.0
–0.5
0.5
0.0
–0.5
B
GHB α-CH2
GHB γ-CH2
GHB β-CH2
9.0
8.5
8.0
7.5
7.0
6.5
6.0
5.5
5.0
4.5
4.0 ppm
3.5
3.0
2.5
2.0
1.5
1.0
FIGURE B2.4.3.5 1H NMR spectra at 600.13 MHz of (A) human saliva (pH 6.87) and of (B) human saliva after the addition of 3.25 r 10 3 mol/dm3 GHB. Abbreviations: A acetate, Ala alanine, Chol choline, Form formate, Gly glycine, His histidine, Lac lactate, lys lysine, NAG N-acetylsugars, Phe phenylalanine, Prop propionate, PUFA polyunsaturated fatty acids, Pyr pyruvate, Suc Succinate, TAG triacylglycerol, TMAO trimethylamine oxide, Tyr tyrosine, Tyr` tyrosine residues, TSP sodium trimethylsilyl propionate [83].
© 2009 by Taylor & Francis Group, LLC
104
The Virtopsy Approach
of ingestion of ethanol in forensic cases [86]. This contrasts with acetaldehyde, the primary metabolite of ethanol from oxidative metabolism in the liver, the blood concentrations of which are difficult to interpret because of nonenzymatic conversion of ethanol to acetaldehyde after sampling [85]. Ethyl glucuronide can be detected in urine for 13−20 h (after drinking small amounts) up to about 3.5 days (after large amounts) [87,88]. Urine as a sample for the analysis of ethyl glucuronide is advisable because the concentration is always much higher than in blood. Unlike ethanol, the urinary concentration of ethyl glucuronide is highly influenced by urine dilution; thus the concentration of ethyl glucuronide should be related to the creatinine content in the sample. In addition to ethyl glucuronide, ethyl sulphate—another non-oxidative metabolite of ethanol—is proposed for measurement in urine [89]. Its temporal window of detection is similar to ethyl glucuronide, but it is formed via a different metabolic pathway. This is why it was proposed to use them conjointly, thereby increasing sensitivity for detecting recent alcohol consumption. In vivo MRS has been used for the measurement of alcohol in the human brain after alcohol consumption. But though 1H spectroscopic measures of brain alcohol were found to be highly correlated with both breath and blood alcohol concentrations after equilibration in brain tissue, some experiments give rise to the hypothesis that there could be more than one compartment of brain alcohol characterized by different molecular environments, suggesting that only a fraction of brain alcohol is visible with 1H MRS [90]. In addition, for reliable quantitation of alcohol concentration with 1H MRS, the effect of relaxation and the subsequent signal decay during acquisition has to be corrected for.
B2.4.3.9 Identification
B2.4.3.8 Potential Applications of NMR Spectroscopy to Diseased Persons
B2.4.3.10.1 Temperature
In deceased persons, in situ MR spectroscopy offers new possibilities, particularly for the systematic and focused investigation of postmortem tissue changes. It is the only method allowing noninvasive and repetitive analysis of multiple metabolites simultaneously, offering an insight into “life” after death. MRS could become a useful tool for the detection of disturbances on a metabolic and functional level, which to date is rather a diagnosis by exclusion. Nevertheless, care needs to be taken concerning the strategy of the study, since the sensitivity of MRS, depending on the tissue and substances of interest, is limited. For an optimal analysis the possibility to measure different nuclei, e.g., 1H, 13C, 31P should be exploited. The advantages of high resolution NMR lie mainly, as discussed in the applications for living persons, in the quick overview of the compounds contained in a sample which, in the case of homogenous body fluids, e.g., serum, CSF, or urine, does not need any time-consuming preparation. Additionally, quantitation of substances is relatively simple and straightforward. For the analysis of blood or tissue samples, an adequate extraction is needed unless MAS for the direct investigation of solids is available.
© 2009 by Taylor & Francis Group, LLC
Identification with, at least, determination of sex, age, and stature is a prerequisite in forensic medicine [91]. As long as the soft tissues of the body are preserved, morphology can be used to determine sex and body stature and to estimate age. In advanced decomposition stages, sex is reliably determined by the means of DNA, while this is not always possible for skeletal remains depending on ambient factors [92]. Concerning the estimation of age, the evaluation of the dental morphology including histological features and the analysis of classic anthropologic parameters stand in the first place [93–95]. An objective and very accurate method is the determination of the degree of aspartic acid racemization (AAR) in dentine or other tissues, e.g., cartilage, ligaments, or bone, which is based on the accumulation of the D-form of the amino acid [96–99]. However, the methodology for the determination of AAR is quite demanding and the preparation of the sample needs several steps depending on the kind of tissue used. For decomposing bodies and well preserved skeletons the determination of age by evaluation of the fat fraction of lumbar bone marrow by MRS has been discussed in detail previously. Of course, the methodology would have to be adapted to the different conditions compared to living persons concerning tissue temperature, water content, and other tissue characteristics. On the other hand, examination time and correction for patient movement is not an issue any longer.
B2.4.3.10 Application of MR Spectroscopy and High Resolution NMR to the Estimation of Postmortem Intervals (PMI)
The relevance of the time of death in forensic medicine is obvious and shall not be discussed here. Methods for the estimation of the time since death based on body temperature— alone or in combination with nontemperature-based indicators [100]—are reliable in the early postmortem interval and have been investigated thoroughly [101-105]. Measurement of body temperature was mostly performed in the rectum, but for certain time intervals, i.e., up to 10.5 hours postmortem, brain temperature alone or in combination with rectal temperature seems to allow still more precise estimations [106,107]. The determination of temperature of various tissues such as brain, skeletal muscle, and other deeply located soft tissues by 1H MRS has already been well described [108–111]. In Childs et al. brain temperature is derived from the chemical shift difference between signals of water and N-acetyl-aspartate (NAA). Figure B2.4.3.6 illustrates the water and NAA peaks for the individual voxels in 1H MR spectroscopic imaging of the brain and the corresponding temperature map [111]. Different MR thermographic methods allow the determination of tissue temperature in a very short time with a precision of less than 1nC [109]. Particularly in combination with a noninvasive MR imaging examination to answer morphological questions, the determination of temperature could be
Internal Body Documentation
105
A
B
C °C 39.0 38.0 37.0 36.0 35.0
pronounced in myocardium than in limb muscle. Unfortunately, the myocardial ATP concentration is additionally influenced by the cause of death; more precisely, by the degree of anoxia immediately before death, thus obscuring the effect of postmortem time. It could be possible that the ATP level was more valuable as an indicator of anoxia before death, provided that the time of death was already known. A further limitation to the use of ATP as a marker of PMI is the short time slot in which changes in ATP levels can actually be observed. A method proposed for the estimation of PMI in the late postmortem interval is the quantification of metabolites correlated with the decomposition of brain tissue [115–117]. The fact that in situ 1H MR spectroscopy allows the simultaneous measurement of several metabolites was taken advantage of. Thus the final model is based on a statistical model consisting of mathematical functions of the individual metabolites, leading to improved accuracy. The animal model and the statistical procedure used for the calculation of an objective value for the PMI with corresponding confidence limits is discussed in Section D2.2 “Investigation of Decomposing Sheep Brain by Means of 1H-MR-Spectroscopy: An Attempt Toward an Objective PMI Estimation in The Later Postmortem Period.” In vitro NMR spectroscopy was used to analyze adipocere in decomposed bodies to evaluate the degree of decomposition [118,119]. The pattern of fatty acids could possibly give hints on the time since death and on ambient factors influencing decomposition. A noninvasive examination of adipocere by MRS would clearly offer practical advantages compared to other methods including high resolution NMR.
34.0
FIGURE B2.4.3.6 MR spectroscopic imaging (MRSI) of the brain of a healthy volunteer showing (A) water and (B) NAA peaks of the individual MRSI voxels, and (C) the corresponding temperature map [111].
precisely measured by MRS. The determination of the temperature at different locations in the body provides additional information on the cooling conditions at the site where the body is found, allowing a better mathematical modeling of the process of cooling [112]. B2.4.3.10.2 Biochemical Markers Numerous methods have been proposed for the estimation of time since death by means of biochemical markers. Their practical value is up to now only limited, due to the fact that these methods did not meet the needs concerning practicability and accuracy [113]. Some examples showing the range of possibilities concerning the application of high-resolution NMR spectroscopy and MRS shall be mentioned here. One of the first attempts where high resolution 31P NMR spectroscopy was used to follow postmortem changes with a possible long term goal of estimating PMI was a study that investigated the decrease of adenosine-tri-phosphate (ATP) in muscle tissue in the early postmortem interval up to 7 hours after death [114]. As expected, a decrease of the ATP level was found, being more
© 2009 by Taylor & Francis Group, LLC
B2.4.3.10.3 MR Parameters as Indicators for Time Since Death As MR parameters reflect tissue characteristics, they can also be correlated with time postmortem. Even though most of the presented studies were not concerned with the forensic aspect of postmortem tissue changes, the results can nevertheless be used for the problem of time since death. The 1H MR relaxation times T1 and T2 were measured in porcine brain tissue samples up to 90 hours after death or operative excision while being stored at 8nC. The T1 values did not show any time dependence and thus would not be useful as an indicator of PMI, while T2 values decreased slightly during the examination interval [120]. In another study, liver tissue samples were characterized by 1H NMR leading to the result that during the interval of at least 120 min after excision T1 does not change, but T2, being more sensitive to water content and pH, possibly allows detection of early nonspecific tissue alterations reflecting postmortem autolysis [121]. Early tissue alterations during the first 60 minutes after death are shown in tissue samples from the liver, pancreas, kidney, testis, spleen, and brain [122]. Figure B2.4.3.7 illustrates the decrease of T1 relaxation time of rat liver and pancreas, with time after tissue excision depending on storage temperature. Additionally, the influence of storage temperature on relaxation times and on cellular and organellar integrity was
106
The Virtopsy Approach
290
300
4°C
28°C
28°C
250 T1 (msec)
260 T1 (msec)
4°C
270
280
240
230
210
220 40°C
40°C 200
190
180
170 0
20
40 60 80 Elapsed Time (min)
100
120
0
20
40 60 80 Elapsed Time (min)
100
120
FIGURE B2.4.3.7 Relaxation time T1 versus time after excision for rat liver (left) and pancreas (right). The measurement was performed at 40nC and 20 MHz. Tissue samples were kept at 4nC, 28nC, or 40nC. Values are the mean of four rats, with standard deviation bars indicated [122].
assessed with MR and electron microscopy, respectively. In contrast to the results of Moser et al., Baba et al. found that overall T2 changes were smaller than T1 changes in all tissues. B2.4.3.11 Cause of Death B2.4.3.11.1 Diagnosis of the Extent of Hypoxia The cause of death is always associated with a certain degree of hypoxia. Particularly pronounced is hypoxia in death by strangulation, e.g., hanging. pH in muscle tissue can be an indicator of hypoxia. Measurement of pH has been performed in postmortem myocardial tissue by in situ 31P MRS based on the fact that the signal of inorganic phosphate shifts dependent on accumulation of lactic acid [123]. Additionally, phosphorus-containing metabolites, e.g., ATP and ADP, creatine phosphate, and phosphates associated with sugars, were analyzed subject to different causes of death in an animal model. In situ 31P MR spectroscopy has the advantage of being noninvasive and can be applied to intact bodies. An alternative method is high resolution 31P NMR spectroscopy using magic angle spinning (MAS), which allows intact tissue samples to be investigated with increased sensitivity and resolution [124]. Figure B2.4.3.8 shows a comparison of a 1H-decoupled 31P MAS spectrum of a tumorous muscle tissue sample with a 1H-decoupled 31P spectrum of the extract of the same piece of tissue. Although the acquisition of the spectrum of the extract took about 20-fold longer, the two spectra are similar regarding resolution. However, MAS can
© 2009 by Taylor & Francis Group, LLC
not be performed on a standard high resolution spectrometer, but needs corresponding equipment such as an MAS probehead and adequate transmitter and receiver units. B2.4.3.11.2 Biochemical Markers as an Aid to Revealing Causes of Death Associated with Metabolic Changes The investigation of metabolic disturbances responsible for death is very difficult, and clinical references are not applicable, as postmortem changes tend to obscure antemortem distinctive metabolic features in most biofluids rapidly. However, the identification of metabolic causes of death is essential in forensic practice and the practicability and usefulness of many biomarkers in diverse biological fluids has been investigated thoroughly and is provided in good overviews [125,126]. Few examples where the application of high resolution NMR or in situ MR spectroscopy could be useful will be discussed. Postmortem vitreous humor has been analyzed by chemical methods concerning the concentration of D-hydroxybutyrate and other possible markers of ketoacidotic coma in diabetes mellitus [127]. Ketoacidotic coma is a serious complication arising from diabetes, particularly type I, and may be the cause of sudden death. Statistically significant differences in the vitreous humor concentration of D-hydroxybutyrate, glucose and lactate, glucose alone, and fructosamine were found between a group of cases with antemortem diagnosis of diabetes mellitus and a group without medical record of diabetes mellitus. While the sum of glucose and lactate
Internal Body Documentation
107
A
B Pi
Pi
PE
PE GPC GPE
PC
5
4
3 ppm
2
1
PC
5
4
GPC GPE 3 ppm
2
1
FIGURE B2.4.3.8 1H-decoupled 31P spectra acquired at 14 T (242 MHz for 31P): (A) MAS spectrum of a sample from a fibrosarcoma tumor (at 4nC, spinning at 3 kHz, 512 scans, repetition time (TR) = 3.35 s, total acquisition time = 28.6 min). (B) high-resolution spectrum of the extract of the same piece of tissue (5500 scans, TR = 6.35 s, total acquisition time = 585 min) [124].
allowed identification of diabetes with a sensitivity of 74% and a specificity of 78%, D-hydroxybutyrate consecutively serves as a marker for ketoacidotic coma. Thus the simultaneous quantification of these biomarkers, particularly glucose and lactate and D-hydroxybutyrate, by high-resolution NMR spectroscopy could facilitate the detection of sudden death due to ketoacidotic coma. To investigate the correlation of renal insufficiency and diabetes mellitus with amino acid abnormalities in human brain, tissue samples from autopsies were analyzed regarding their content of free amino acids [128]. While there were no significant differences in the amino acid composition related to postmortem interval or to diabetes mellitus, samples from persons with renal disease showed a significant cerebral increase of urea, phenylethanolamine, and I-aminobutyric acid. This result could possibly contribute to the understanding of pathophysiological mechanisms involved in uremic encephalopathy and supplement the postmortem diagnosis of renal insufficiency. A study from the same group has also shown a correlation of specific amino acid abnormalities with severe liver diseases [129]. As such, the analysis of free amino acids in human brain is quite promising as a supplementation to other postmortem investigations in helping to reveal functional importance of diseases. In situ MR spectroscopy of the brain is able to identify a pattern of metabolites, including amino acids, in a single spectrum. As metabolic abnormalities might be concentrated in certain brain regions, localized information on metabolite concentrations can be acquired. As an isolated compartment similar to vitreous humor, synovial fluid from cases without known metabolic disturbances, joint disease, or resuscitation history was analyzed by chemical methods for usual clinical laboratory markers such as electrolytes, amino acids, and glucose [130]. The values were compared to those from vitreous humor. As the
© 2009 by Taylor & Francis Group, LLC
range of the values of the two matrices approximately corresponds, the authors conclude that synovial fluid may be used as a similar tool to vitreous humor, also with the same restrictions. However, the analysis is more difficult because of the higher viscosity, and because of the unknown delay for concentration adjustments between blood and synovial fluid acute metabolic disorders cannot be identified. Though it has been tried to elucidate the relation of changes of biochemical parameters in synovial fluid with causes of death, it remains vague, and pathophysiological explanations are either mostly missing or trivial [131,132]. Thus though the measurement of synovial fluid with high-resolution NMR spectroscopy would be interesting from a methodological point of view, it does not seem to supply sufficient new information. B2.4.3.12 Toxicology B2.4.3.12.1 Drugs in Biological Fluids, Tissue Specimens and Pharmaceutical Samples In forensic toxicology, different matrices such as biofluids, tissue samples, and less frequently also unknown pills have to be analyzed with a need for high quantitative accuracy and precision. High-resolution NMR spectroscopy allows the determination and quantification of drugs in pharmaceutical and urine samples with a high selectivity and resolving power without any preparatory separation steps. An example shows quantitative analysis of antibiotics, i.e., miconazole, metronidazole, and sulfamethoxazole, either separately or in the form of admixtures with detection limits below reported pharmacopeial methods [133]. Statistical tests revealed no significant differences in precision compared to commonly used high performance liquid chromatography (HPLC) methods in this study. In another report 31P NMR spectroscopy was used for quantification of glyphosate, a herbicide,
108
The Virtopsy Approach
in two suicide cases [134]. Blood and urine samples did not require any pretreatment apart from addition of D2O, while liver tissue required an enzymatic digestion before measurement. With a 200-MHz NMR spectrometer, levels of 1 mg/ ml could be detected in less than a minute. History shows that the application of in vitro NMR spectroscopy in the context of identifying unknown drugs in forensic toxicology has already been proposed in 1970 for the identification of barbiturates in tissue specimens, biofluids, and other materials obtained from autopsy [135,136] and for the identification and quantification of ureides, methaqualon, and barbiturates in tissue samples [137]. Apparently sometimes toxicologists had to work with a high grade of creativity, as only the combined application of several different spectroscopic methods was successful. A study from 1973 describes how complementary spectroscopic data obtained from UV-, IR-, NMR-, and mass spectrometry was used for the identification of an unknown tablet ingredient [138]. Later, the use of 13C NMR spectroscopy as a stand-alone technique was proposed for the analysis of a series of compounds used as illicit drugs or in combination with them (e.g., cocaine, amphetamine, phenacetin, etc.). The same study highlights the potential of NMR for analyzing complex forensic drug mixtures with no need for preliminary separation of the components [139]. To date, high resolution NMR spectroscopy is an accepted tool for the investigation of misused drugs in forensic science and for the determination of the purity of reference drug standards and the routine analysis of illicit drugs and adulterants [17,140].
The examination of vitreous humor (VH) by means of NMR spectroscopy has, to our knowledge, not been performed to date, though VH as a relatively simple matrix would be well suited for. Additionally, vitreous humor has the particular advantage that it is much less susceptible to possible postmortem changes compared to blood. In a study of an animal model VH is proposed as a complementary sample to blood for the detection of diazepam [141]. The kinetics of the drug levels in both samples subject to time after administration and postmortem distribution and redistribution phenomena were investigated. Drug analysis was performed by a series of extractions followed by high performance liquid chromatography (HPLC) with different detectors for each kind of sample. Vitreous humor seems to be a particularly useful specimen for the quantification of 6-monoacetylmorphine (6-MAM), which is used to establish the origin of morphine from heroin exposure [142]. As VH seems to be relatively secluded from the blood circulation, certain metabolites such as 6-MAM seem to remain detectable in VH for a much longer interval than in blood, or show higher concentrations. For the analysis of 6-MAM, the samples underwent a series of analytical steps including extraction, derivatization, and gas chromatography-mass spectrometry (GC-MS). The time-consuming and elaborate procedure in the examples of diazepam and 6-MAM, though part of standard analytical procedures in forensic toxicology laboratories, raises the idea of using NMR spectroscopy, since no sample pretreatment is necessary. NMR has already been successfully applied for the determination of, for example, diazepam and 6-MAM [140]. Figure B2.4.3.9 shows the aromatic region of a proton spectrum
LPA
B
B
P AH
C 06 06
H
16.22 9.57 8.1
8.0
2.76 7.9
16.39 7.8
7.7
7.6
7.5
37.61
7.4 7.3 7.2 Chemical Shift (ppm)
1.18 7.1
7.0
8.57 6.9
N
1.87 1.76 6.8
6.7
1.08 63.11 6.6
6.5
FIGURE B2.4.3.9 Aromatic region of the 1H NMR spectrum of illicit heroin. Abbreviations: A acetaminophen, B benzocaine, C caffeine, H heroin, L lidocaine, N noscapine, O6 O6-monoacetylmorphine, P procaine [140].
© 2009 by Taylor & Francis Group, LLC
Internal Body Documentation
109
100 90 80
Percent Heroin HCl
70 60
NMR CE GC
50 40 30 20 10 0 0
20
40
60
80 100 120 Sample Number
140
160
180
200
FIGURE B2.4.3.10 Analyses of 196 heroin samples by NMR, capillary electrophoresis (CE), and gas chromatography (GC) showing good agreement between the methods [140].
of illicit heroin in D2O, which in this case consisted of eight different substances including 6-MAM. Every compound except acetaminophen had at least one non-overlapping signal, allowing simple quantification. In Figure B2.4.3.10, the results of three different analytical methods, i.e., NMR, gas chromatography (GC), and capillary electrophoresis (CE) for the analysis of heroin samples are shown. They are in good agreement. B2.4.3.12.2 GHB in Postmortem Toxicology As previously mentioned, gamma-hydroxybutyrate (GHB) is generally present in body fluids of healthy persons from endogenous production. The discrimination from exposure is thus an important issue for forensic toxicologists. This concern is increased in postmortem toxicology since additional influences such as the postmortem interval or storage temperature have to be considered [143]. A correlation of postmortem production in liver tissue with time after death has been reported [144], and also under laboratory conditions the comparison of GHB concentration of unpreserved with sodium fluoride-preserved samples after extended postmortem intervals supports usage of sample preservation [145]. The analysis and interpretation of concentrations seems to be problematic, depending on the analytical methods used and the type of biological fluids examined (blood, urine, bile, vitreous humor). However, although interpretative cutoff levels under which detected GHB could be endogenous in nature for blood and urine are proposed [145], clear and systematic investigations are missing. As GHB is easily detected and quantified with 1H NMR spectroscopy in biofluids [83], the use of 1H NMR spectroscopy may be most promising for elucidating problems related to the analysis and interpretation of GHB concentrations.
© 2009 by Taylor & Francis Group, LLC
B2.4.3.12.3 Postmortem Markers of Chronic Alcoholism and Premortem Ingestion of Ethanol The postmortem diagnosis of chronic alcoholism is not obvious in forensic medicine, where knowledge of medical history is often missing. Nevertheless, chronic alcoholism is not only very common in the forensic case population, but is also associated with characteristic pathology and causes of death. In addition, the traditional method of diagnosing chronic alcoholism with evaluation of blood alcohol concentration and liver histology is rather non-specific, while the analysis of hair can be used only with a certain length of hair and is quite time-consuming. Thus a facilitated diagnosis by measurement of biomarkers would be desirable. In an evaluation of the diagnostic value of I-glutamyltransferase (GGT) and carbohydrate-deficient transferrin (CDT) in comparison to the standard parameters alcoholic liver disease (ALD), blood alcohol concentration (BAC), presence of multiple bruises, and poor hygiene of the feet, Sadler et al. concluded that the measurement of serum CDT was the most promising marker [146]. In fact, a large amount of scientific literature has been written on CDT as a biomarker of chronic alcohol abuse in both clinical and forensic environment, pointing at different aspects, i.e., structure of CDT and the pathomechanism of its ethanol induced increase, analytical methods including electrophoretic, chromatographic, immunometric, and massspectrometric methods [147–150], as well as evaluation of its diagnostic value and factors influencing the data interpretation [151,152]. Basically, carbohydrate-deficient transferrin is an isoform of the serum iron-transporting protein composed of 679 amino acids usually containing carbohydrate side chains with up to four terminal sialic acid residues. Associated with
110
alcohol abuse is primarily a fraction called asialo-transferrin, which in healthy humans can only be determined in traces. Factors influencing the postmortem concentration have been evaluated from a forensic point of view [153,154]. While the concentration of CDT was not significantly altered for at least 15 days of storage at 4nC and 20nC, respectively, and no correlation was found with the site of blood collection (femoral/cardiac), repeated freezing and thawing as well as hemolysis of the sample resulted in decreased levels of CDT. In 25 consecutive forensic autopsy cases, the sensitivity and specificity of identifying antemortem alcohol abuse by the serum concentration of CDT was found to be 88% and 63%, respectively, applying a self-established cutoff level for positive alcohol abuse [153]. Improved sensitivity (95%) and specificity (71%) were found for the determination of CDT in vitreous humor of forensic autopsy cases [155] using the detection limit of the test (5 U/l) as a cut-off level. However, although vitreous CDT is stable during laboratory handling, the authors could not exclude the fact that time-dependent changes of vitreous total transferrin concentration could possibly affect CDT during a prolonged postmortem interval [156]. To overcome the problems and specific deficiencies of using any single marker, Brinkmann et al. proposed the application of a combination of four different markers of alcoholism in serum, i.e., methanol, acetone and 2-propanol, I-glutamyl-transferase (GGT), and CDT in an application to blood samples from living persons [157]. A formula for a so-called “Alc-Index” combining the markers with different mathematical weights facilitated the definition of a threshold between alcoholics and nonalcoholics, resulting in a differentiation of the two groups with 93% sensitivity and 100% specificity. These values are quite appealing, suggesting that this concept could possibly also be applied to postmortem samples. However, even though for each marker different and specifically adapted chemical methods exist, the simultaneous quantification of several markers by high resolution NMR would rigorously simplify the procedure, particularly because no sample pretreatment besides solving in a suitable deuterated solvent is needed. To date and to our knowledge, the use of NMR spectroscopy for the postmortem detection of biomarkers for chronic alcohol abuse has not yet been documented. Without postmortem sampling of body fluids, the diagnosis of premortem ethanol ingestion can be based for at least 12 hours after death on the analysis of fatty acid ethyl esters (FAEE) in liver and adipose tissue [158,159]. FAEE are cytotoxic nonoxidative metabolites of ethanol synthesized primarily in organs commonly damaged by ethanol abuse such as the liver, pancreas, heart, and brain [160]. At least for plasma and serum samples, FAEE levels are influenced by gender and storage at room temperature for more than 24 hours, and the correlation between blood ethanol and FAEE was improved when accounting for triglycerides [161]. It was also demonstrated that chronic alcohol abuse is correlated with a high concentration of a specific FAEE species, i.e., ethyl oleate, helping to differentiate binge drinkers from chronic alcoholics [162]. High resolution NMR spectroscopy
© 2009 by Taylor & Francis Group, LLC
The Virtopsy Approach
studies exist investigating mainly the binding properties of FAEE [163,164]. To avoid time-consuming and elaborate preanalytical extraction procedures, in situ 13C MR spectroscopy for the analysis of FAEE in adipose tissue would perfectly meet the needs. In vivo analysis of lipid composition of human adipose tissue has been reported using different MRS techniques leading to improved separation of individual resonances and better specificity regarding the identification of the individual components of the fatty acids [165–167]. Further technical and methodological improvements are expected. The analysis of adipose tissue by in situ MRS in forensic cases seems to be promising. B2.4.3.13 Conclusion NMR spectroscopy provides information on chemical composition and metabolism either in vitro or in situ. Though the use of NMR spectroscopy in forensic toxicology was proposed about 40 years ago, the ideas about the application of this method in research and routine diagnostics have almost been lost again. Since analytical NMR in chemistry and physics as well as MRI and MRS in medicine and biology have seen an incredible development in the past two decades, some ideas about pre- and postmortem applications of NMR in the forensic field should be revisited. This chapter discusses potential applications of NMR spectroscopy in situ and in vitro. When examinations are presented that are done with non-NMR methods today, it does not necessarily mean that they can or should be replaced by NMR spectroscopy or MRS. Potential applications are presented without the claim of being reasonable or even feasible, but with the intention to create a consciousness about these opportunities in the light of modern NMR and MRS with drastically improved sensitivity and resolution as compared to the early days. This may motivate forensic scientists to tackle the many unsolved problems and questions in forensic science using modern methods that have proven their enormous value in chemistry and medicine.
B2.4.3.14 References 1. Boesch, C. Molecular aspects of magnetic resonance imaging and spectroscopy, Mol. Aspects. Med., 20, 185, 1999. 2. Boesch, C. Magnetic Resonance Spectroscopy: Basic Principles, in Clinical Magnetic Resonance Imaging, Edelman, R.R. et al., Eds., Saunders, Elsevier, Philadelphia, PA , 2005, 459. 3. De Graaf, R.A. In vivo NMR spectroscopy: Principles and techniques, John Wiley & Sons, Chichester UK, 1999, 1. 4. Gadian, D.G. Nuclear magnetic resonance and its applications to living systems, Clarendon Press, Oxford, 1982, 1. 5. Friebolin, H. Ein- und zweidimensionale NMR-Spektroskopie: Eine Einführung, 3rd, Wiley-VCH, Weinheim, D, 1999, 1. 6. Bigler, P. NMR Spectroscopy: Processing Strategies Spectroscopic Techniques: An Interactive Course, 2, Wiley-VCH, Weinheim, 2000, 1. 7. Chang, C., Chen, G.C., and Jang, T., A critical assessment of brain metabolites: analysis of perchloric acid extracts using proton nuclear magnetic resonance, Neurosci. Lett., 196, 134, 1995.
Internal Body Documentation
8. Burri, R. et al., Brain development: 1H magnetic resonance spectroscopy of rat brain extracts compared with chromatographic methods, Neurochem. Res., 15, 1009, 1990. 9. Fan, T.W. et al., Combined use of 1H-NMR and GC-MS for metabolite monitoring and in vivo 1H-NMR assignments, Biochim. Biophys. Acta, 882, 154, 1986. 10. Gribbestad, I.S. et al., 1H NMR spectroscopic characterization of perchloric acid extracts from breast carcinomas and non-involved breast tissue, NMR Biomed., 7, 181, 1994. 11. Edzes, H.T. et al., Analysis of phospholipids in brain tissue by 31P NMR at different compositions of the solvent system chloroform-methanol-water, Magn Reson. Med., 26, 46, 1992. 12. Le Belle, J.E. et al., A comparison of cell and tissue extraction techniques using high-resolution 1H-NMR spectroscopy, NMR Biomed., 15, 37, 2002. 13. Tyagi, R.K. et al., Simultaneous extraction of cellular lipids and water-soluble metabolites: evaluation by NMR spectroscopy, Magn Reson. Med., 35, 194, 1996. 14. Henke, J. et al., Combined extraction techniques of tumour cells and lipid/phospholipid assignment by two dimensional NMR spectroscopy, Anticancer Res., 16, 1417, 1996. 15. Ross, B.D. and Danielsen, E.R., Magnetic resonance spectroscopy diagnosis of neurological diseases, Marcel Dekker, New York, 1999, 1. 16. Howe, F. et al., Proton spectroscopy in vivo, Magnetic Resonance Quarterly, 9, 31, 1993. 17. Groombridge, C.J. NMR spectroscopy in forensic science, Annual Reports on NMR Spectroscopy, 32, 215, 1996. 18. Ettaro, L., Berger, R.P., and Songer, T., Abusive head trauma in young children: characteristics and medical charges in a hospitalized population, Child Abuse Negl., 28, 1099, 2004. 19. Keenan, H.T. et al., A population-based study of inflicted traumatic brain injury in young children, JAMA, 290, 621, 2003. 20. Ludwig, S. and Warman, M., Shaken baby syndrome: a review of 20 cases, Ann. Emerg. Med., 13, 104, 1984. 21. King, W.J., MacKay, M., and Sirnick, A., Shaken baby syndrome in Canada: clinical characteristics and outcomes of hospital cases, CMAJ., 168, 155, 2003. 22. Berger, R.P. et al., Assessment of the macrophage marker quinolinic acid in cerebrospinal fluid after pediatric traumatic brain injury: insight into the timing and severity of injury in child abuse, J. Neurotrauma, 21, 1123, 2004. 23. Berger, R.P., Kochanek, P.M., and Pierce, M.C., Biochemical markers of brain injury: could they be used as diagnostic adjuncts in cases of inflicted traumatic brain injury?, Child Abuse Negl., 28, 739, 2004. 24. Berger, R.P. et al., Identification of inflicted traumatic brain injury in well-appearing infants using serum and cerebrospinal markers: a possible screening tool, Pediatrics, 117, 325, 2006. 25. Berger, R.P. and Kochanek, P.M., Urinary S100B concentrations are increased after brain injury in children: A preliminary study*, Pediatr. Crit Care Med., 7, 557, 2006. 26. Satchell, M.A. et al., Cytochrome c, a biomarker of apoptosis, is increased in cerebrospinal fluid from infants with inflicted brain injury from child abuse, J. Cereb. Blood Flow Metab, 25, 919, 2005. 27. Sinz, E.H. et al., Quinolinic acid is increased in CSF and associated with mortality after traumatic brain injury in humans, J. Cereb. Blood Flow Metab, 18, 610, 1998. 28. Bell, M.J. et al., Quinolinic acid in the cerebrospinal fluid of children after traumatic brain injury, Crit Care Med., 27, 493, 1999.
© 2009 by Taylor & Francis Group, LLC
111
29. Berger, R.P. et al., Serum neuron-specific enolase, S100B, and myelin basic protein concentrations after inflicted and noninflicted traumatic brain injury in children, J. Neurosurg., 103, 61, 2005. 30. Berger, R.P. et al., Serum biomarkers after traumatic and hypoxemic brain injuries: insight into the biochemical response of the pediatric brain to inflicted brain injury, Dev. Neurosci., 28, 327, 2006. 31. Thomas, D.G.T., Palfreyman, J.W., and Ratcliffe, J.G., Serum-myelin-basic-protein assay in diagnosis and prognosis of patients with head injury, Lancet, 1, 113, 1978. 32. Parizel, P.M. et al., Cortical hypoxic-ischemic brain damage in shaken-baby (shaken impact) syndrome: value of diffusion-weighted MRI, Pediatr. Radiol., 33, 868, 2003. 33. Biousse, V. et al., Diffusion-weighted magnetic resonance imaging in Shaken Baby Syndrome, Am. J. Ophthalmol., 133, 249, 2002. 34. Morad, Y. et al., Shaken baby syndrome without intracranial hemorrhage on initial computed tomography, J. AAPOS., 8, 521, 2004. 35. Morad, Y. et al., Normal computerized tomography of brain in children with shaken baby syndrome, J. AAPOS., 8, 445, 2004. 36. Blumenthal, I. Shaken baby syndrome, Postgrad. Med. J., 78, 732, 2002. 37. Sato, Y. et al., Head injury in child abuse: evaluation with MR imaging, Radiology, 173, 653, 1989. 38. American Academy of Pediatrics: Committee on Child Abuse and Neglect and Committee on Children With Disabilities. Assessment of maltreatment of children with disabilities, Pediatrics, 108, 508, 2001. 39. Duhaime, A.C. et al., Nonaccidental head injury in infants: the “shaken-baby syndrome,” N. Engl. J. Med., 338, 1822, 1998. 40. Adelson, P.D. et al., Cerebrovascular response in infants and young children following severe traumatic brain injury: a preliminary report, Pediatr. Neurosurg., 26, 200, 1997. 41. Lee, Y. et al., MR imaging of shaken baby syndrome manifested as chronic subdural hematoma, Korean J. Radiol., 2, 171, 2001. 42. Beaulieu, C. et al., Diffusion-weighted magnetic resonance imaging: theory and potential applications to child neurology, Semin. Pediatr. Neurol., 6, 87, 1999. 43. Liu, A.Y. et al., Traumatic brain injury: diffusion-weighted MR imaging findings, AJNR Am. J. Neuroradiol., 20, 1636, 1999. 44. Cowan, F.M. et al., Early detection of cerebral infarction and hypoxic ischemic encephalopathy in neonates using diffusion-weighted magnetic resonance imaging, Neuropediatrics, 25, 172, 1994. 45. Haseler, L.J. et al., Evidence from proton magnetic resonance spectroscopy for a metabolic cascade of neuronal damage in shaken baby syndrome, Pediatrics, 99, 4, 1997. 46. Kreis, R. et al., Hypoxic encephalopathy after near-drowning studied by quantitative 1H-magnetic resonance spectroscopy, J. Clin. Invest, 97, 1142, 1996. 47. Ross, B.D., Ernst, T., and Kreis, R., Proton magnetic resonance spectroscopy in hypoxic ischemic disorders, in MRI of the Central Nervous System in Infants and Children, Bax, M. and Faerber, E.N., Eds., MacKeith Press, London, 1996, 48. Holshouser, B.A. et al., Proton MR spectroscopy in children with acute brain injury: comparison of short and long echo time acquisitions, J. Magn Reson. Imaging, 11, 9, 2000.
112
49. Ashwal, S. et al., Predictive value of proton magnetic resonance spectroscopy in pediatric closed head injury, Pediatr. Neurol., 23, 114, 2000. 50. Makoroff, K.L. et al., Elevated lactate as an early marker of brain injury in inflicted traumatic brain injury, Pediatr. Radiol., 35, 668, 2005. 51. Schmeling, A. et al., Age estimation, Forensic Sci. Int., 165, 178, 2007. 52. Ritz-Timme, S. et al., Empfehlungen für die Altersdiagnostik bei Lebenden im Rentenverfahren, Rechtsmed., 12, 193, 2002. 53. Schmeling, A. et al., Studies on the time frame for ossification of the medial clavicular epiphyseal cartilage in conventional radiography, Int. J. Legal Med., 118, 5, 2004. 54. Lockemann, U. et al., Arbeitsgemeinschaft für Forensische Altersdiagnostik der Deutschen Gesellschaft für Rechtsmedizin - Empfehlungen für die Altersdiagnostik bei Jugendlichen und jungen Erwachsenen ausserhalb des Strafverfahrens, Rechtsmed., 14, 123, 2004. 55. Schmeling, A. et al., Age estimation of unaccompanied minors. Part I. General considerations, Forensic Sci. Int., 159 Suppl 1, S61-S64, 2006. 56. Garamendi, P.M. et al., Reliability of the methods applied to assess age minority in living subjects around 18 years old. A survey on a Moroccan origin population, Forensic Sci. Int., 154, 3, 2005. 57. Olze, A. et al., Combined determination of selected radiological and morphological variables relevant for dental age estimation of young adults, Homo., 56, 133, 2005. 58. Maber, M., Liversidge, H.M., and Hector, M.P., Accuracy of age estimation of radiographic methods using developing teeth, Forensic Sci. Int., 159 Suppl 1, S68-S73, 2006. 59. Willems, G. et al., Dental age estimation in Belgian children: Demirjian’s technique revisited, J. Forensic Sci., 46, 893, 2001. 60. Kugel, H. et al., Age- and sex-specific differences in the 1H-spectrum of vertebral bone marrow, J. Magn Reson. Imaging, 13, 263, 2001. 61. Schellinger, D. et al., Normal lumbar vertebrae: anatomic, age, and sex variance in subjects at proton MR spectroscopy--initial experience, Radiology, 215, 910, 2000. 62. Schellinger, D. et al., Potential value of vertebral proton MR spectroscopy in determining bone weakness, AJNR Am. J. Neuroradiol., 22, 1620, 2001. 63. De Bisschop, E. et al., Fat fraction of lumbar bone marrow using in vivo proton nuclear magnetic resonance spectroscopy, Bone, 14, 133, 1993. 64. Yeung, D.K. et al., Osteoporosis is associated with increased marrow fat content and decreased marrow fat unsaturation: a proton MR spectroscopy study, J. Magn Reson. Imaging, 22, 279, 2005. 65. Rucker, G. Die qualitative Analyse von Barbitursäurederivaten durch Kernresonanzspektroskopie, Arch. Pharm., 299, 688, 1966. 66. Imbenotte, M. et al., Identification and quantitation of xenobiotics by 1H NMR spectroscopy in poisoning cases, Forensic Sci. Int., 133, 132, 2003. 67. Azaroual, N. et al., Valproic acid intoxication identified by 1H and 1H-(13)C correlated NMR spectroscopy of urine samples, MAGMA., 10, 177, 2000. 68. Cartigny, B. et al., 1H NMR spectroscopic investigation of serum and urine in a case of acute tetrahydrofuran poisoning, J. Anal. Toxicol., 25, 270, 2001. 69. Wahl, A. et al., Poisoning with methanol and ethylene glycol: 1H NMR spectroscopy as an effective clinical tool for diagnosis and quantification, Toxicology, 128, 73, 1998.
© 2009 by Taylor & Francis Group, LLC
The Virtopsy Approach
70. Janus, T. et al., H nuclear magnetic resonance spectroscopic investigation of urine for diagnostic and clinical assessment of methanol intoxication, Basic Clin. Pharmacol. Toxicol., 97, 257, 2005. 71. Imbenotte, M. et al., Detection and quantitation of xenobiotics in biological fluids by 1H NMR spectroscopy, J. Toxicol. Clin. Toxicol., 41, 955, 2003. 72. Maschke, S. et al., Salicylate poisoning: two-dimensional J-resolved NMR urinalysis, NMR Biomed., 8, 19, 1995. 73. Maschke, S. et al., Diagnosis of a case of acute chloroquine poisoning using 1H NMR spectroscopy: characterisation of drug metabolites in urine, NMR Biomed., 10, 277, 1997. 74. Maschke, S. et al., Detection by 1H-NMR spectroscopy of chloroquine in urine from acutely poisoned patient, Clin. Chem., 43, 698, 1997. 75. Cartigny, B. et al., Determination of glyphosate in biological fluids by 1H and 31P NMR spectroscopy, Forensic Sci. Int., 143, 141, 2004. 76. Cartigny, B. et al., 1H NMR urine analysis as an effective tool to detect creatine supplementation, J. Anal. Toxicol., 26, 355, 2002. 77. Cartigny, B. et al., Analysis of ingested material and urine by GC-MS and 1H NMR spectroscopy: poisoning of an adult with adulterated soda, J. Anal. Toxicol., 30, 86, 2006. 78. Brenneisen, R. et al., Pharmacokinetics and excretion of gamma-hydroxybutyrate (GHB) in healthy subjects, J. Anal. Toxicol., 28, 625, 2004. 79. Scharf, M.B. et al., Pharmacokinetics of gammahydroxybutyrate (GHB) in narcoleptic patients, Sleep, 21, 507, 1998. 80. Thai, D. et al., Gamma-hydroxybutyrate and ethanol effects and interactions in humans, J. Clin. Psychopharmacol., 26, 524, 2006. 81. Yeatman, D.T. and Reid, K., A study of urinary endogenous gamma-hydroxybutyrate (GHB) levels, J. Anal. Toxicol., 27, 40, 2003. 82. Crookes, C.E. et al., A reference range for endogenous gamma-hydroxybutyrate in urine by gas chromatographymass spectrometry, J. Anal. Toxicol., 28, 644, 2004. 83. Grootveld, M. et al., Determination of the illicit drug gamma-hydroxybutyrate (GHB) in human saliva and beverages by 1H NMR analysis, Biofactors, 27, 121, 2006. 84. Bales, J.R. et al., Use of high-resolution proton nuclear magnetic resonance spectroscopy for rapid multi-component analysis of urine, Clin. Chem., 30, 426, 1984. 85. Jones, A.W. Urine as a biological specimen for forensic analysis of alcohol and variability in the urine-to-blood relationship, Toxicol. Rev., 25, 15, 2006. 86. Hoiseth, G. et al., A pharmacokinetic study of ethyl glucuronide in blood and urine: Applications to forensic toxicology, Forensic Sci. Int., 2007. 87. Stephanson, N. et al., Direct quantification of ethyl glucuronide in clinical urine samples by liquid chromatography-mass spectrometry, Ther. Drug Monit., 24, 645, 2002. 88. Wurst, F.M. et al., Ethyl glucuronide discloses recent covert alcohol use not detected by standard testing in forensic psychiatric inpatients, Alcohol Clin. Exp. Res., 27, 471, 2003. 89. Wurst, F.M. et al., Ethyl sulphate: a direct ethanol metabolite reflecting recent alcohol consumption, Addiction, 101, 204, 2006. 90. Fein, G. and Meyerhoff, D.J., Ethanol in human brain by magnetic resonance spectroscopy: correlation with blood and breath levels, relaxation, and magnetization transfer, Alcohol Clin. Exp. Res., 24, 1227, 2000. 91. Rosing, F.W. et al., Recommendations for the forensic diagnosis of sex and age from skeletons, Homo., 58, 75, 2007.
Internal Body Documentation
92. von Wurmb-Schwark, N. et al., Extraction and amplification of nuclear and mitochondrial DNA from ancient and artificially aged bones, Leg. Med. (Tokyo), 5 Suppl 1, S169-S172, 2003. 93. Cattaneo, C. Forensic anthropology: developments of a classical discipline in the new millennium, Forensic Sci. Int., 165, 185, 2007. 94. Rissech, C. and Malgosa, A., Pubis growth study: Applicability in sexual and age diagnostic, Forensic Sci. Int., 2007. 95. Martrille, L. et al., Comparison of four skeletal methods for the estimation of age at death on white and black adults, J. Forensic Sci., 52, 302, 2007. 96. Ritz-Timme, S. et al., Age estimation: the state of the art in relation to the specific demands of forensic practise, Int. J. Legal Med., 113, 129, 2000. 97. Ritz-Timme, S. and Collins, M.J., Racemization of aspartic acid in human proteins, Ageing Res. Rev., 1, 43, 2002. 98. Ritz-Timme, S., Laumeier, I., and Collins, M., Age estimation based on aspartic acid racemization in elastin from the yellow ligaments, Int. J. Legal Med., 117, 96, 2003. 99. Mornstad, H., Pfeiffer, H., and Teivens, A., Estimation of dental age using HPLC-technique to determine the degree of aspartic acid racemization, J. Forensic Sci., 39, 1425, 1994. 100. Henssge, C. et al., Experiences with a compound method for estimating the time since death. II. Integration of nontemperature-based methods, Int. J. Legal Med., 113, 320, 2000. 101. Henssge, C. et al., Experiences with a compound method for estimating the time since death. I. Rectal temperature nomogram for time since death, Int. J. Legal Med., 113, 303, 2000. 102. Marshall, T.K. and Hoare, F.D., Estimating the time of death: The rectal cooling after death and its mathematical expression, J. Forensic Sci., 7, 56, 1962. 103. Shapiro, H.A. The post-mortem temperature plateau, J. Forensic Med., 12, 137, 1965. 104. Marty, W. Thermographie und Thermometrie in der Forensik mit besonderer Berücksichtigung der Todeszeitbestimmung, Habil.Schrift, Zürich, 1995, 105. Henssge, C. and Madea, B., Estimation of the time since death in the early post-mortem period, Forensic Sci. Int., 144, 167, 2004. 106. Henssge, C. et al., [Determination of the time of death by measurement of central brain temperature], Z. Rechtsmed., 93, 1, 1984. 107. Henssge, C. et al., [Determination of the time of death based on simultaneous measurement of brain and rectal temperatures], Z. Rechtsmed., 93, 123, 1984. 108. Corbett, R., Laptook, A., and Weatherall, P., Noninvasive measurements of human brain temperature using volumelocalized proton magnetic resonance spectroscopy, J. Cereb. Blood Flow Metab, 17, 363, 1997. 109. Wlodarczyk, W. et al., Comparison of four magnetic resonance methods for mapping small temperature changes, Phys. Med. Biol., 44, 607, 1999. 110. Yoshioka, Y. et al., Noninvasive measurement of temperature and fractional dissociation of imidazole in human lower leg muscles using 1H-nuclear magnetic resonance spectroscopy, J. Appl. Physiol, 98, 282, 2005. 111. Childs, C. et al., Determination of regional brain temperature using proton magnetic resonance spectroscopy to assess brain-body temperature differences in healthy human subjects, Magn Reson. Med., 57, 59, 2007. 112. Mall, G. and Eisenmenger, W., Estimation of time since death by heat-flow Finite-Element model. Part I: method, model, calibration and validation, Leg. Med. (Tokyo), 7, 1, 2005.
© 2009 by Taylor & Francis Group, LLC
113
113. Madea, B. Is there recent progress in the estimation of the postmortem interval by means of thanatochemistry?, Forensic Sci. Int., 151, 139, 2005. 114. Gothoda, M. and Harada, H., Application of nuclear magnetic resonance (NMR) spectroscopy to legal medicine|ATP quantitation in tissue with 31P-NMR, Tokushima J. Exp. Med., 40, 61, 1993. 115. Ith, M. et al., Observation and identification of metabolites emerging during postmortem decomposition of brain tissue by means of in situ 1H-magnetic resonance spectroscopy, Magn. Reson. Med., 48, 915, 2002. 116. Scheurer, E. et al., Statistical evaluation of time-dependent metabolite concentrations: estimation of post-mortem intervals based on in situ 1H-MRS of the brain, NMR Biomed., 18, 163, 2005. 117. Banaschak, S. et al., Estimation of postmortem metabolic changes in porcine brain tissue using 1H-MR spectroscopypreliminary results, Int. J. Legal Med., 119, 77, 2005. 118. Szathmary, S.C., Von Tamaska, L., and Steigel, A., [Postmortem decomposition of neutral lipids. Use of modern methods of analysis (HPLC, capillary GC, GC-MS and NMR) in adipocere formation], Z. Rechtsmed., 94, 273, 1985. 119. Takatori, T. and Yamaoka, A., Separation and identification of 9-chloro-10-methoxy (9-methoxy-10-chloro)hexadecanoic and octadecanoic acids in adipocere, Forensic Sci. Int., 14, 63, 1979. 120. Gyorffy-Wagner, Z. et al., Proton magnetic resonance relaxation times T1 and T2 related to postmortem interval. An investigation on porcine brain tissue, Acta Radiol. Diagn. (Stockh), 27, 115, 1986. 121. Moser, E., Holzmueller, P., and Gomiscek, G., Liver tissue characterization by in vitro NMR: tissue handling and biological variation, Magn Reson. Med., 24, 213, 1992. 122. Baba, Y. et al., Time after excision and temperature alter ex vivo tissue relaxation time measurements, J. Magn Reson. Imaging, 4, 647, 1994. 123. Akaiwa, T., Harada, H., and Maeiwa, M., Post-mortem muscular changes studied by nuclear magnetic resonance (NMR) spectroscopy--studies on the 31P-NMR spectrum of the myocardium after sudden death, Nippon Hoigaku Zasshi, 42, 363, 1988. 124. Payne, G.S. et al., Evaluation of 31P high-resolution magic angle spinning of intact tissue samples, NMR Biomed., 19, 593, 2006. 125. Coe, J.I. Postmortem chemistry: practical considerations and a review of the literature, J. Forensic Sci., 19, 13, 1974. 126. Kernbach-Wighton, G. Möglichkeiten postmortal-biochemischer Diagnostik, Rechtsmed., 16, 27, 2006. 127. Osuna, E. et al., Postmortem vitreous humor beta-hydroxybutyrate: its utility for the postmortem interpretation of diabetes mellitus, Forensic Sci. Int., 153, 189, 2005. 128. Schmidt, P. et al., Brain amino acid abnormalities in kidney disease and diabetes mellitus, Forensic Sci. Int., 142, 221, 2004. 129. Musshoff, F. et al., Brain amino acid abnormalities in liver disease|a postmortem study, J. Forensic Sci., 48, 1379, 2003. 130. Madea, B., Kreuser, C., and Banaschak, S., Postmortem biochemical examination of synovial fluid–a preliminary study, Forensic Sci. Int., 118, 29, 2001. 131. More, D.S. and Arroyo, M.C., Biochemical changes of the synovial liquid in corpses with regard to the cause of death. 1: Calcium, inorganic phosphorus, glucose, cholesterol, urea nitrogen, uric acid, proteins, and albumin, J. Forensic Sci., 30, 541, 1985.
114
132. More, D.S. and Arroyo, M.C., Biochemical changes of the synovial liquid of corpses with regard to the cause of death. 2: Alkaline phosphatase, lactic acid dehydrogenase (LDH), and glutamic oxalacetic transaminase (GOT), J. Forensic Sci., 30, 547, 1985. 133. Salem, A.A., Mossa, H.A., and Barsoum, B.N., Application of nuclear magnetic resonance spectroscopy for quantitative analysis of miconazole, metronidazole and sulfamethoxazole in pharmaceutical and urine samples, J. Pharm. Biomed. Anal., 41, 654, 2006. 134. Dickson, S.J. et al., Rapid determination of glyphosate in postmortem specimens using 31P NMR, J. Anal. Toxicol., 12, 284, 1988. 135. Lackner, H. and Doring, G., [NMR spectroscopy identification of barbiturates. I. Efficiency of the method], Arch. Toxikol., 26, 220, 1970. 136. Doring, G. and Lackner, H., [NMR spectroscopy identification of barbiturates. II. Studies on biological material], Arch. Toxikol., 26, 237, 1970. 137. Rucker, G., Bohn, G., and Fell, A.F., Identification and quantitative determination of ureides, methaqualone and barbiturates in autopsy material by NMR-spectroscopy, Arch. Toxikol., 27, 168, 1971. 138. Bohn, G. and Rucker, G., Combined use of spectroscopic methods in forensic analysis as demonstrated on the example of identifying an unknown drug in tablets, Beitr. Gerichtl. Med., 30, 12, 1973. 139. Alm, S. et al., The use of 13C-NMR spectroscopy in forensic drug analysis, Forensic Sci. Int., 19, 271, 1982. 140. Hays, P.A. Proton nuclear magnetic resonance spectroscopy (NMR) methods for determining the purity of reference drug standards and illicit forensic drug seizures, J. Forensic Sci., 50, 1342, 2005. 141. Teixeira, H.M. et al., Vitreous humour as a complementary sample to blood for the detection/confirmation of diazepam: ante-mortem and post-mortem studies in an animal model, Hum. Exp. Toxicol., 23, 571, 2004. 142. Wyman, J. and Bultman, S., Postmortem distribution of heroin metabolites in femoral blood, liver, cerebrospinal fluid, and vitreous humor, J. Anal. Toxicol., 28, 260, 2004. 143. Kintz, P. et al., GHB in postmortem toxicology. Discrimination between endogenous production from exposure using multiple specimens, Forensic Sci. Int., 143, 177, 2004. 144. Sakurada, K. et al., Production of gamma-hydroxybutyric acid in postmortem liver increases with time after death, Toxicol. Lett., 129, 207, 2002. 145. Elliott, S.P. Further evidence for the presence of GHB in postmortem biological fluid: implications for the interpretation of findings, J. Anal. Toxicol., 28, 20, 2004. 146. Sadler, D.W., Girela, E., and Pounder, D.J., Post mortem markers of chronic alcoholism, Forensic Sci. Int., 82, 153, 1996. 147. Tagliaro, F. et al., Carbohydrate-deficient transferrin determination revisited with capillary electrophoresis: a new biochemical marker of chronic alcohol abuse, J. Capill. Electrophor. Microchip. Technol., 6, 137, 1999. 148. Wuyts, B. and Delanghe, J.R., The analysis of carbohydrate-deficient transferrin, marker of chronic alcoholism, using capillary electrophoresis, Clin. Chem. Lab Med., 41, 739, 2003.
© 2009 by Taylor & Francis Group, LLC
The Virtopsy Approach
149. Legros, F.J. et al., Carbohydrate-deficient transferrin isoforms measured by capillary zone electrophoresis for detection of alcohol abuse, Clin. Chem., 48, 2177, 2002. 150. Martello, S. et al., Determination of carbohydrate deficient transferrin (CDT) with capillary electrophoresis: an inter laboratory comparison, Forensic Sci. Int., 141, 153, 2004. 151. Sorvajarvi, K. et al., Sensitivity and specificity of carbohydratedeficient transferrin as a marker of alcohol abuse are significantly influenced by alterations in serum transferrin: comparison of two methods, Alcohol Clin. Exp. Res., 20, 449, 1996. 152. Bortolotti, F., De, P.G., and Tagliaro, F., Carbohydratedeficient transferrin (CDT) as a marker of alcohol abuse: a critical review of the literature 2001-2005, J. Chromatogr. B Analyt. Technol. Biomed. Life Sci., 841, 96, 2006. 153. Malcolm, R. et al., Carbohydrate-deficient transferrin and alcohol use in medical examiner cases, Alcohol, 17, 7, 1999. 154. Simonnet, C. et al., Review of factors susceptible of influencing post-mortem carbohydrate-deficient transferrin, Forensic Sci. Int., 106, 7, 1999. 155. Berkowicz, A. et al., Carbohydrate-deficient transferrin in vitreous humour: a marker of possible withdrawal-related death in alcoholics, Alcohol Alcohol, 36, 231, 2001. 156. Berkowicz, A. et al., Analysis of carbohydrate-deficient transferrin (CDT) in vitreous humour as a forensic tool for detection of alcohol misuse, Forensic Sci. Int., 137, 119, 2003. 157. Brinkmann, B. et al., ROC analysis of alcoholism markers-100% specificity, Int. J. Legal Med., 113, 293, 2000. 158. Refaai, M.A. et al., Ethyl arachidonate is the predominant fatty acid ethyl ester in the brains of alcohol-intoxicated subjects at autopsy, Lipids, 38, 269, 2003. 159. Salem, R.O. et al., Fatty acid ethyl esters in liver and adipose tissues as postmortem markers for ethanol intake, Clin. Chem., 47, 722, 2001. 160. Soderberg, B.L. and Laposata, M., Fatty acid ethyl esters: markers of ethanol intake, Am. Clin. Lab, 20, 18, 2001. 161. Soderberg, B.L. et al., Preanalytical variables affecting the quantification of fatty acid ethyl esters in plasma and serum samples, Clin. Chem., 45, 2183, 1999. 162. Laposata, M. et al., Fatty acid ethyl esters: recent observations, Prostaglandins Leukot. Essent. Fatty Acids, 67, 193, 2002. 163. Cistola, D.P. and Small, D.M., Fatty acid distribution in systems modeling the normal and diabetic human circulation. A 13C nuclear magnetic resonance study, J. Clin. Invest, 87, 1431, 1991. 164. Bird, D.A., Laposata, M., and Hamilton, J.A., Binding of ethyl oleate to low density lipoprotein, phospholipid vesicles, and albumin: a 13C NMR study, J. Lipid Res., 37, 1449, 1996. 165. Dimand, R.J. et al., Adipose tissue abnormalities in cystic fibrosis: noninvasive determination of mono- and polyunsaturated fatty acids by carbon-13 topical magnetic resonance spectroscopy, Pediatr. Res., 24, 243, 1988. 166. Thomas, E.L. et al., An in vivo 13C magnetic resonance spectroscopic study of the relationship between diet and adipose tissue composition, Lipids, 31, 145, 1996. 167. Hwang, J.H. et al., In vivo characterization of fatty acids in human adipose tissue using natural abundance 1H decoupled 13C MRS at 1.5 T: clinical applications to dietary therapy, NMR Biomed., 16, 160, 2003.
B3
3D Visualization of Radiological Data
CONTENTS B3.1 Visualization of Radiological Data: Basics ....................................................................................................................115 B3.1.1 Introduction .......................................................................................................................................................115 B3.1.2 Equipment..........................................................................................................................................................115 B3.1.3 Reconstruction Techniques................................................................................................................................116 B3.1.3.1 Multiplanar Reconstruction (MPR) ..................................................................................................116 B2.1.3.2 Maximum Intensity Projection (MIP) ..............................................................................................116 B3.1.4 Minimum Intensity Projection (MinIP) ............................................................................................................116 B3.1.4.1 Volume Editing ................................................................................................................................118 B3.1.4.2 Shaded Surface Display (SSD) ........................................................................................................ 120 B3.1.4.3 Volume-Rendering Technique (VRT) ............................................................................................. 121 B3.1.5 Further Reading ................................................................................................................................................ 121 B3.2 3D Visualization of Radiological Data: Future ............................................................................................................. 121 B3.2.1 Introduction ...................................................................................................................................................... 121 B3.2.2 The Future of Virtual Autopsies....................................................................................................................... 122 B3.2.2.1 Data Acquisition .............................................................................................................................. 122 B3.2.2.2 Data Management ............................................................................................................................ 124 B3.2.2.3 Data Visualization ........................................................................................................................... 125 B3.2.2.4 Virtual Autopsy Workstation ........................................................................................................... 127 B3.2.3 References......................................................................................................................................................... 128
B3.1 VISUALIZATION OF RADIOLOGICAL DATA: BASICS Steffen Ross and Michael J. Thali
B3.1.1 INTRODUCTION The introduction of spiral/helical computed tomography (CT) in the early 1990s led to exciting new applications in the field of three-dimensional reconstruction of the former, mostly two-dimensional, axial CT cross-sections. 3D-reconstructed CT data enable the depiction of a variety of pathologies at a glance, even from a radiologically inexperienced viewpoint (e.g., police officers, attorneys, judges). Spiral data acquisition allows—in contrast to the serial, sequential data acquisition—the registration of a whole-volume dataset without misregistration of anatomic detail. The three-dimensional images integrate the information of a large stack of two-dimensional images, which are often easier to interpret than the original cross-sections (Figure B3.1.1). For a better understanding it may be helpful to think of a volume dataset as a digital cube, floating within the computer monitor and organized in subelements called voxels (similar to the pixels in 2D pictures). Isotropy of a dataset, which means cuboid voxels, allows the reconstruction of a 3D model in all three axes (x,y,z) (Figure B3.1.2) without a loss of spatial resolution.
Each 3D rendering technique is based on mathematical formulas, determining what portion (voxel) of the data volume is displayed on the two-dimensional computer screen and how that portion should be weighted for the best representation of anatomic relationships. Overall, there are three main rendering techniques actually used in reconstruction of CT data: 1. Maximum intensity projection (MIP) and minimum intensity projection (MinIP) 2. Shaded surface display (SSD) 3. Volume-rendering technique (VRT)
B3.1.2 EQUIPMENT Due to the great performance progress within the last two decades, computer technology is rapidly making issues of reconstruction time and storage capacity less significant. Today it is possible to build a reconstruction workstation with common off-the-shelf components. There are many vendors with a variety of workstation and software packages on the market (Figure B3.1.3). For the purposes of forensic imaging, the purchase of a fully equipped system is not mandatory. For instance, there is temporarily no need to perform perfusion studies or high-resolution imaging of the (still uncontrasted) coronary arteries. If one does not wish to use commercially 115
© 2009 by Taylor & Francis Group, LLC
116
The Virtopsy Approach
FIGURE B3.1.1 Left side: representative axial slices of a head CT; right side: VRT reconstruction of the whole dataset. (Suicidal gunshot; point of entry right parietal; exit point at the vertex.)
available systems, it is possible to build a homemade workstation with free software relatively inexpensively. The most important thing to consider is usability of a system, performance, and, last but not least, scalability.
B3.1.3 RECONSTRUCTION TECHNIQUES B3.1.3.1 Multiplanar Reconstruction (MPR) Multiplanar reconstruction is the simplest method of reconstruction. As mentioned in the introduction, volume is built by stacking the axial slices. The reconstruction software then cuts slices through the volume in a different plane (usually orthogonal). MPR is frequently used for examining the spine. Axial images through the spine will only show one vertebral
body at a time and cannot reliably show compression fractures along the longitudinal axis of the body (Figure B3.1.4). By reformatting the volume, it becomes much easier to visualize the structure, and the position of one vertebral body in relation to the others. Curved-plane reconstruction can be performed on the volume dataset. This allows bent or curved structures, like the sacrum, to be “straightened” so that the entire length can be visualized in one image (Figure B3.1.5). B2.1.3.2 Maximum Intensity Projection (MIP) The concept of MIP is the evaluation of a volume by drawing a virtual light ray from the observer’s eye through the data volume and selecting the voxel with the highest attenuation (maximum intensity; Figure B3.1.6). This voxel is then used as the displayed pixel of the rendered picture. MIP is quite useful in creating angiographic pictures from CT and magnetic resonance imaging (MRI) data. Although postmortem angiography is one use for this forensic imaging routine, there are other useful applications for this method, such as visualization of a misguided catheter or a coronary stent, which can be easily done with MIP (Figure B3.1.7 and Figure B3.1.8). This method also has a number of principle-related shortcomings. As mentioned, only the material with the highest intensity along the projected light ray will be represented in the reconstructed picture, obscuring adjacent structures in the volume. This limitation can be partially overcome with the use of volume editing in the form of section-by-section editing (slab view) or interactively (sliding slab).
B3.1.4 MINIMUM INTENSITY PROJECTION (MINIP) FIGURE B3.1.2 Model of a 3D dataset with isotropic (cuboid) voxels.
© 2009 by Taylor & Francis Group, LLC
The concept of MinIP is similar to MIP, but in this method the voxel with the least attenuation along a virtual light ray is visualized. MinIP is the method of choice for visualizing
3D Visualization of Radiological Data
117
FIGURE B3.1.3 Leornardo workstation (Siemens Medical Solutions, Erlangen, Germany) used in the Virtopsy Group for 3D reconstructions and data analysis.
FIGURE B3.1.4 Multiplanar reconstruction (MPR) of the thoracolumbar spine with the corresponding 3D reconstruction of the volume (VRT) in the lower right corner. Compression/bursting fractures of the eighth thoracal vertebral body and the first lumbar vertebral body. The compression and the lateral misplacement on level Th8 is much better depicted in the sagittal and coronal reconstructions than in the axial view.
© 2009 by Taylor & Francis Group, LLC
118
The Virtopsy Approach
FIGURE B3.1.5 The curved reconstruction allows a complete visualization of the bended sacrum in one picture. Note the extended fracture of the right sacral wing.
structures with very low x-ray attenuation, such as air or other gases, seen in soft-tissue emphysema, pneumothorax, air embolism, or lung lacerations (Figure B3.1.9 and Figure B3.1.10).
FIGURE B3.1.7 MIP reconstruction of a whole-body MSCT scan. Notice the flipped jugular catheter in the inferior vena cava with catheter tip in the right atrium. Additional visualization of a tracheal tubus, the tip of a subclavian venous catheter, and a chest tube from the left.
B3.1.4.1 Volume Editing Editing a radiological dataset is common in forensic imaging to remove other high-attenuation structures such as the CT
FIGURE B3.1.6 Scheme of virtual light ray falling through a row of voxels with different x-ray attenuations. The voxel with the highest attenuation (the “whitest”) will be displayed in the MIP reconstruction.
© 2009 by Taylor & Francis Group, LLC
FIGURE B3.1.8 MIP reconstruction of a stent in the right coronary, used for identification purposes in a decomposed corpse.
3D Visualization of Radiological Data
119
FIGURE B3.1.9 Coronal MinIP of the thorax of a polytraumatized traffic accident victim. Notice the gas embolism in the left heart ventricle/ascending thoracic aorta (*), the ruptured left diaphragm (red arrows), the right-sided apical pneumothorax (white arrow), and the soft-tissue emphysema (black arrow).
table or foreign bodies from the incident scene, such as items that might have been left in the body bag. The two options for manipulating a 3D dataset are manual and automated editing. Manual editing involves the user drawing regions of interest around the structures to be included or excluded from the 3D picture. On fast workstations with real-time volume rendering this operation does not necessarily need to be done on the axial source images; it can also be done by cutting unwanted details out of the actual 3D image (Figure B3.1.11).
FIGURE B3.1.10 MinIP reconstruction of the torso of decomposed body. The intravascular gas and the gas in the soft tissue are well depicted.
Automated editing includes the segmentation of a dataset by differentiation of certain tissue by their individual attenuation. Until now this has not been a routine task because of the overlapping attenuation values of different tissue and the
FIGURE B3.1.11 Postmortem cranial MSCT. Interactive volume editing is used to remove unwanted structures, such as remains of the body bag and the partly scanned collarbones, with the aim of providing a free view of the skull.
© 2009 by Taylor & Francis Group, LLC
120
The Virtopsy Approach
FIGURE B3.1.12 Surface rendering of a skull with a frontal impression fracture on the left. On the right a VRT of the same dataset from nearly the same perspective.
inability of the software to distinguish them from each other. This task is a subject of ongoing research. B3.1.4.2 Shaded Surface Display (SSD) SSD, or surface rendering, was with its development in the early 1970s the first rendering technique applied to medical datasets. SSD is a rendering style in which surfaces are
defined by a threshold value. Voxel intensities defined within the threshold value appear as a surface in the 3D picture (Figure B3.1.12). Surface contours are typically modeled as a number of overlapping polygons. By computing a virtual light source for each polygon, it is possible to display the object as a shaded surface model. By setting partial opacities for certain predefined surfaces, it is possible to show
FIGURE B3.1.13 VRT of a whole-body MSCT with different voxel opacities for skin, muscle, and bone.
© 2009 by Taylor & Francis Group, LLC
3D Visualization of Radiological Data
anatomic structures with different intensities in one reconstruction (e.g., skin and bone together). Advantages of SSD are high reconstruction speed and flexibility in image rendering. Small bony details, such as structures of the skull base or cranial sutures, are well defined in SSD. A drawback of this method is that the surface is derived from only a small percentage of the available data (less than 10% in most cases). SSD is not adequate for structures with no well-differentiated surfaces. For instance, undislocated fracture systems with narrow bony fissures will vanish in an SSD because of the poor differentiation between the adjacent fragments and smoothing artifacts of the extracted surface data.
121
view through an actual endoscope. This fascinating method of traveling through body cavities is limited due to the segmental flattening of the colon and obstructed airways. The incorporation of information from the whole volume can lead to a higher fidelity of displayed data. Handling and display of these large datasets in near-real-time is a job for high-speed double- or quad-processor workstations with large amounts of random access memory (RAM) and hard-disk space. Computing performance is fundamental for minimal waiting time and maximal interactivity with the 3D model. Differences in the integration of various volume-rendering algorithms result in varying quality and usability among the great number of applications available on the market.
B3.1.4.3 Volume-Rendering Technique (VRT) As the name implies, VRT renders the whole volume and not only surfaces defined by threshold values. VRT takes the entire volume data and sums the contributions of each voxel along a line from the viewer’s eye through the 3D dataset. The resulting composite of each voxel is then displayed on the screen. Defining a window with and level, like in conventional axial CT images, makes it possible to segment certain tissue (e.g., skin, muscle, and bone). A transfer function maps the input data values to lighting properties needed by volume-rendering algorithms. These properties are opacity, brightness, and color. A recent 3D workstation provides a wide variety of VRT presets, making a tissue-specific reconstruction simply a matter of a few mouse clicks (Figure B3.1.13). Special uses for VRT are virtual bronchioscopy and colonoscopy (Figure B3.1.14). In these modes a virtual camera is placed inside an air-filled body cavity, which provides an endoscopic view similar to the
B3.1.5 FURTHER READING Calhoun, P. S., B. S. Kuszyk, D. G. Heath, J. C. Carley, and E. K. Fishman. 1999. Three-dimensional volume rendering of spiral CT data: theory and method. Radiographics 19(3): 745–64. Choplin, R. H., J. M. Farber, K. A. Buckwalter, and S. Swan. 2004. Three-dimensional volume rendering of the tendons of the ankle and foot. Seminars in Musculoskeletal Radiology 8(2): 175–83 (June). Pavone, P., G. Luccichenti, and F. Cademartiri. 2001. From maximum intensity projection to volume rendering. Seminars in Ultrasound, CT, and MR 22(5): 413–19 (October). Rodt, T., S. O. Bartling, J. E. Zajaczek, M. A. Vafa, T. Kapapa, O. Majdani, et al. 2006. Evaluation of surface and volume rendering in 3D-CT of facial fractures. Dento Maxillofacial Radiology 35(4): 227–31 (July).
B3.2 3D VISUALIZATION OF RADIOLOGICAL DATA: FUTURE Anders Persson, Anders Ynnerman, Claes Lundström, and Patric Ljung
B3.2.1 INTRODUCTION
FIGURE B3.1.14 Example of virtual colonoscopy, view in the gasfilled colon transversum. Notice the typical colonic haustration.
© 2009 by Taylor & Francis Group, LLC
This section uses the current state of the art in data acquisition and visualization of radiological data as its starting point. Current trends and challenges for the future development of virtual autopsies (VAs) are described, forming the basis for specifications of research agendas for development of the underlying methods and extraction of the knowledge needed to fully exploit the potential of VAs. The reader of this text is assumed to have a working knowledge of existing methods for 3D visualization in medicine such as described in the previous chapters of this book. It has been shown that the VA distinctively better exploits its value when combined with recent direct volume-rendering (DVR) techniques to handle the imaging data of the bodies. The overall goal of the VA procedure is that it can be applied to every victim of any kind of incident. This goal is, however, far from being reached, and several challenges remain to be addressed before the fully envisioned potential of the VA can be realized. The needed advances encompass both technology and workflow issues in VA procedures.
122
The Virtopsy Approach
The foundation of the development of VAs is the modern imaging modalities that can generate large, high-quality datasets with submillimeter precision. Interactive visualization of these 3D datasets can provide valuable insight into the corpses, and enables noninvasive diagnostic procedures. Efficient handling and analysis of these datasets is, however, problematic. For instance, in postmortem CT imaging, not being limited by a certain patient depending radiation dose the datasets can, however, be generated with such a high resolution that they become difficult to handle in today’s archive retrieval and interactive visualization systems, specifically in the case of full-body scans. Herein we describe the current trends in the areas of data acquisition, data archiving, data visualization, and data analysis that will enable the next generation of VA tools. The development of these technologies must take place in close synergy with research on VA workflows and case categories; for example, victims of airplane crashes, bombings, or flooding require novel visualization techniques. Technical and methodological advances for the VA must be unified to reach its full potential from the forensic perspective.
B3.2.2 THE FUTURE OF VIRTUAL AUTOPSIES B3.2.2.1 Data Acquisition B3.2.2.1.1 Plain Film Imaging Because of the low soft-tissue discrimination and the complex anatomy of the body, it is easy to miss soft-tissue pathology, bleedings, and small foreign artifacts with conventional postmortem imaging. Dual-energy imaging can be used to overcome this problem. Two approaches are common: In the first approach a copper filter separates two phosphor plates. The first plate captures the full-energy spectrum, but the second captures only the higher-energy spectrum. The combination is then used to generate dual-energy soft-tissue and bone images. One standard exposure thus creates a conventional image and the two dual-energy images, with no additional radiation. The second approach to dual-energy imaging is sequential exposure, used with newer flat-panel detectors. Not enough radiation passes through to perform two-plate image capture, so two images are obtained in rapid sequence: one at a low kV level and the second at a high kV level. B3.2.2.1.2 CT B3.2.2.1.2.1 Multienergy CT Multienergy data can be acquired in several different ways with CT: r r r r
Using energy discrimination detectors Switching kVP in x-ray tube Scanning the corpse twice with different energy Using a dual-source CT scanner with two x-ray tubes
Dual-energy CT (DECT) with two x-ray sources running simultaneously at different energies can acquire two datasets showing different attenuation levels. DECT allows obtaining
© 2009 by Taylor & Francis Group, LLC
additional information about the elementary chemical composition of a CT-scanned material. The explanation for this is the different relative strengths of the photoabsorption process, which is the dominant attenuation process at low photon energies, and Compton scattering, which prevails at larger photon energies [1]. The largest strengths are met for hydrogen with weak photoabsorption and for contrast agents containing iodine with strong photoabsorption. The relative strength of photoabsorption concerning Compton scattering can be determined by using two different average photo energies, which corresponds to two different tube voltages (80 and 140 kV). In conclusion, x-ray absorption is energy dependent; for example, scanning an object with 80 kV results in a different attenuation than with 140 kV. This physics phenomenon can help to discriminate materials of an atomic number that are very similar to one another, such as calcium and iodine contrast, by using changes in CT numbers between the two energy settings. Color can then be assigned according to CT the value changes between 80 and 140 kV, and a colormapped, dual-energy image can differentiate between calcifications and iodine contrast. This technique can also be used for direct subtraction of bone from the CT raw data without postprocessing. Conventional bone removal approaches require a high degree of user interaction or often suffer from artifacts due to motion between the unenhanced and the enhanced acquisition. This technique can also be used to better visualize blood clots in vessels and sometimes bleedings in soft tissue [2]. The material-specific difference in attenuation shown in the resulting image could facilitate classifications of different tissue types such as tendons and cartilage, and could make CT a more effective VA tool. Further research is needed to explore this new technique (Figure B3.2.1). B3.2.2.1.2.2 Multienergy Spectral CT In the future we most likely will use multienergy spectral CT modalities for VAs. New detectors can count each individual x-ray photon. This leads to improved tissue classification and datasets with higher resolution. Spectral CT will further stress the visualization pipeline. B3.2.2.1.2.3 Volumetric Inverse Geometry CT (IGCT) Another example of a new CT concept is the volumetric inverse-geometry CT (IGCT). A conventional CT scanner has a point source of x-rays and a large detector array, and simultaneously collects projection data along all the ray paths connecting the point source with each detector element, called a “forward geometry.” In an “inverse-geometry” CT system, there are many sources separated from each other and a much smaller detector array [3,4]. As in a forward-geometry system, the source and detector arrays rotate around the object. The inverse geometry has a number of potential advantages. If sources are distributed in both the axial and transverse directions, the system can have arbitrary axial coverage while being immune from the cone-beam artifacts that cone-beam systems based on a single point source suffer.
3D Visualization of Radiological Data
123
CMIV
CMIV
FIGURE B3.2.1 DECT with two x-ray sources running simultaneously at different energies allows obtaining additional information about the elementary chemical composition of computer tomography scanned material. (Left) Tendons and small vessels can be visualized without IV contrast. (Right) Ligaments between the carpal bones are visualized.
Another significant challenge in IGCT is the need to obtain a large in-plane field of view (FOV). The FOV is principally determined by the size of the source array, and constructing a very large (e.g., 80 cm) scanned source array would be technically challenging. Two solutions to this challenge are being investigated. One uses a smaller source array but multiple detector arrays separated in the transverse direction. The other uses a large array of discrete sources. IGCT has the potential to acquire energy resolved datasets in high resolution. B3.2.2.1.2.4 C-Arm Flat Panel CT This type of modality is currently used for classic angiography, especially for neuroinvasive procedures. It can produce low-quality CT scans and 3D images. In the future this type of modality may play an important role in the field of VAs. B3.2.2.1.3 Magnetic Resonance Imaging (MRI) B3.2.2.1.3.1 Synthetic MRI Using MRI as a means of virtual autopsy might pose difficulties in the generation of good contrast images. The
A
B
temperature of the body influences the MR relaxation times of all tissue, and hence clinically established protocols need to be adjusted for optimal contrast at any given temperature. A better approach is the direct measurement of the absolute MR tissue parameters for tissue characterization, in this case longitudinal Tl relaxation, transverse T2 relaxation, and proton density (PD). The quantification methods have become faster in recent years and can be achieved within a reasonable time [5–7] (Figure B3.2.2). Since quantification data may be difficult to interpret, the approach of synthetic MRI can be used. Synthetic MRI means that the three absolute parameters T1, T2, and PD are translated into ordinary MR contrast images [8,9]. This can be done by calculating the expected image intensity based on the absolute tissue parameters and a particular choice of scanner settings, such as echo time TE, repetition time TR, and flip angle A. The scanner parameters for the synthesis are free to choose, and hence an unlimited range of T1- or T2-weighted contrast images can be generated based on a single quantification scan. Optimization of contrast as a function of temperature can be achieved in a postprocessing
C
FIGURE B3.2.2 Example of the resulting images of a quantification measurement on the brain of a patient with multiple sclerosis. A single axial slice is shown out of the 22 that were acquired in a scan time of 5:18 minutes. The in-plane resolution is 1 mm2; the thickness is 3 mm. (A) Absolute T1 relaxation (scale 0–2000 ms). (B) Absolute T2 relaxation (scale 0–300 ms). (C) Absolute proton density (scale 500–1000, where 1000 corresponds to pure water at 37n).
© 2009 by Taylor & Francis Group, LLC
124
The Virtopsy Approach
A
B
C
D
E
F
FIGURE B3.2.3 Contrast images of ordinary MRI (A–C) compared with images of synthetic contrast MRI (D–F) on the same patient as in Figure B3.2.2. The ordinary images are made separately whereas the synthetic images all come from the same dataset. The total scan time of both methods is comparable. A single axial slice is shown out of the 22 that were acquired. The in-plane resolution is 1 mm2; the thickness is 3 mm. (A,D) T1-weighted image with TE 15 ms, TR 595 ms, flip angle 69n. (B,E) T2-weighted image with TE 100 ms, TR 4450 ms, flip angle 90n. (C,F) T2-weighted FLAIR image with TE 120 ms, TR 6000 ms, flip angle 90n, IR delay 2000 ms.
step. In fact, the approach can be reversed such that the absolute value of the MR parameters of a reference tissue (e.g., serum) may indicate the actual temperature inside the body (Figure B3.2.3). Additionally, synthetic MRI may generate stronger, nonphysical, contrast images that cannot be acquired directly on the scanner. The voxels of the acquired data can be looked upon as coordinates in a T1-T2-PD space where each tissue forms a three-dimensional cluster. These clusters can be visualized separately showing a single tissue of interest. An alternative approach is to relate the T1-T2-PD combination to a red, blue, and green (RBG) color scale such that each tissue acquires a specific color composition depending on its MR tissue parameters. This results in a visual segmentation of tissue. Since the MR parameters are absolute, an identical color transformation will lead to a specific color-to-tissue relation, and deviations are easier to discover (Figure B3.2.4). B3.2.2.2 Data Management B3.2.2.2.1 Managing Large Datasets The increased capacity of imaging modalities, which result in volumetric high-resolution datasets, and the potential of multivariate and fused datasets pose significant challenges for medical data archiving systems (i.e., Picture Archiving and Communication Systems [PACS]). The challenge is furthermore underscored by the fact that departments and facilities in the workflow may be at physically distributed locations, all
© 2009 by Taylor & Francis Group, LLC
requiring access to data at common storage systems. Access patterns to data depend on the type of visualization, ranging from overview to detail and combining different data fields in the same visual representation. Although the capacity of modern computers increases the gap between processing power and data access, latency and transfer bandwidth is widening. For the visualization and analysis of these large datasets it therefore becomes necessary for the archiving systems to provide data at different resolution levels and different data fields, depending on the task to be performed. The virtual autopsy serves as an excellent test bed for such new technologies, and recent advances in the visualization community provide a basis for forthcoming archiving and visualization systems. A major change of paradigm is to consider the 3D volumetric data rather than stacks of 2D image slices, and in addition to provide access to data at different resolution levels. There is a tremendous gain that can be made in both data loading and visualization performance if data resolution is optimized against resulting visual quality in volume rendering. In the literature, such techniques are referred to as multiresolution level-of-detail techniques. Some of the most recent techniques employ adaptive level of detail depending on the currently used transfer function, which depicts the visible content of a volume (Figure B3.2.5). Technical descriptions are found in [10–12]. By using a multiresolution approach it is possible to visualize and analyze datasets at the high visual quality and resolution originally provided by modern imaging devices on standard desktop PCs equipped with advanced commodity
3D Visualization of Radiological Data
A
125
B
C
FIGURE B3.2.4 Examples of synthetic vector imaging, an alternative visualization method on the same patient as the previous two figures. The MR tissue parameters are taken as coordinates in a T1-T2-PD space The contrast in the image is based on the vector length in this space between the tissue. The origin of the space can be placed inside a single tissue cluster of coordinates and is displayed as white. Subsequently, the vector length to all other tissue is calculated and displayed as a gray scale decreasing to black. (A) CSF enhancement by displaying the distance from the nominal CSF values of T1 4200 ms, T2 2000 ms, and PD 1000. (B) White-matter enhancement by displaying the distance from the nominal white-matter values of T1 630 ms, T2 75 ms, and PD 650. The specific value for black in the contrast image can also be specified. (C) White corresponds to MS plaque at values of T1 1300 ms, T2 150 ms, and PD 850; black corresponds to white matter at values of T1 630 ms, T2 75 ms, and PD 650. The gray-scale values are determined by the ratio RWM/RWM-MS, where RWM is the distance from the coordinate to nominal white matter and RWM-MS is the distance from nominal MS plaque to nominal white matter.
graphics cards. Using such a system, one may start viewing the full dataset and quickly obtain an overview, and then zoom in on specific regions of interest of the body. In PACS, viewing parameters|most importantly the transfer function|are evaluated against the content of the volume. By dividing the volume into small blocks, about a cubic centimeter, each block can be assigned a different resolution so the overall visual quality is optimized while the requested data size is limited to the available memory capacity of the graphics card. When a specific part of the body is zoomed in on, all blocks with any visible content are usually assigned the full resolution and quality of the original dataset. B3.2.2.3 Data Visualization B3.2.2.3.1 A DVR Pipeline for VAs Use of volume rendering is becoming more widespread in the medical community, and visualization methods have rapidly matured. There are, however, many visually oriented tasks
that are specific to VAs, and there is potential for development of tailored methods addressing the specific needs of these tasks. Initial work in this direction has recently taken place, and in this section a selected subset of several new technologies and rendering techniques is presented that has specific bearing on VAs. B3.2.2.3.1.1 Detecting Alien Objects The virtual autopsy procedure enables fast detection and localization of metal fragments, such as bullets and splinters. Metal fragments are highly absorbent of x-ray radiation and are therefore easily detected on the intensity scale of CT images. It is, however, useful to render images with context, showing bones and relevant soft tissue as well. A consequence is then that metal fragments appear as buried in the data due to CT reconstruction artifacts, the partial volume effect, and occlusion by other visible body parts. This is illustrated by the left image in Figure B3.2.6, which is created using traditional DVR.
FIGURE B3.2.5 A multiresolution system provides data at different resolution levels that enables interactive viewing of the whole body as well as detailed zoom-ins of specific region of interest. Adaptive level of detail also provides a significant data reduction by using reduced resolution for parts not visible.
© 2009 by Taylor & Francis Group, LLC
126
The Virtopsy Approach
FIGURE B3.2.6 In basic volume rendering it is hard to discover and find small metal fragments buried in the body, as shown in the left image. Dual TF DVR uses a secondary TF to detect metal fragments that are superimposed onto the final rendered image, providing an “x-ray” rendering, shown in the right image. In this example even fragments at the size of almost a single voxel can be detected, as seen in the inset image to the right.
By specifying that a specific component of the transfer function indicates a material of special interest (metals), it can be captured in the rendering process and superimposed over a traditional image. This is efficiently achieved in the right image of Figure B3.2.6 using advanced, yet commodity, programmable graphics hardware. Alternative approaches also exist; one technique uses preprocessing of the data to detect voxels in the vicinity of metal fragments. By using additional derived attributes, tissue and materials of interest can be further highlighted in the rendering. Figure B3.2.7 shows the same dataset as in Figure B3.2.6 but rendered using a multivariate dataset of derived attributes
from preprocessing of the volume data, a technique previously presented in [13]. B3.2.2.3.1.2 Visualization of Gas Volume rendering of cadavers also delivers improved support for detecting pockets of gases within the body. Gas, air, or postmortem gases can be rendered and hence made perceptible. Gas within the body in places it should not be may provide important clues in forensic investigations. In the right image of Figure B3.2.8, pockets of air are detected in the soft tissue of the chest cavity, to the right of the trachea, indicating that air has been permitted through the
FIGURE B3.2.7 Volume rendering using multivariate volume data. Additional volume attributes are computed in a preprocessing stage and enable the volume renderer to pick up visible content in multidimensional transfer functions. In this case the vicinity of high-density voxels (i.e., voxels containing metal fragments) is highlighted to efficiently show the location of bullet fragments.
© 2009 by Taylor & Francis Group, LLC
3D Visualization of Radiological Data
127
FIGURE B3.2.8 In DVR it is possible to visualize pockets of gases, in this case air, inside the body. During a laryngoscopy procedure this 3-week-old infant died. The virtual autopsy study of the CT scan suggested that air had been permitted around the heart. A ventilation needle had penetrated the pericardium in an attempt to vent out air from the lung cavity, thus allowing air in the pericardium and likely causing the heart to stop. Pockets of air are detected in the soft tissue of the chest cavity to the right of the trachea, the inset in the right image.
pericardium, and likely caused the heart to stop. Rendering images with an emphasis on the air–tissue interface reveals such important findings. B3.2.2.3.1.3 Novel Shading and Illumination Techniques Most current volume-rendering approaches use shading techniques based on models developed for surface-based rendering, typically polygon surface geometry. In volume rendering, a derived surface normal is approximated by the direction of the gradient of the data. This model is arguably appropriate for samples where the gradient magnitude is large. Volume rendering, however, frequently involves samples having a small gradient magnitude. In addition, the gradient is commonly computed with a small support of 3 × 3 × 3 voxels at most. Thus, the immediate surrounding 3D structure has little or no impact on the shading of voxels and therefore constitutes an inadequate model to convey 3D structure. By using a larger area and taking into account the classified samples (i.e., after mapping the volume data through the transfer function), improved perception of the 3D structure is possible. Recent methods again exploit the data reduction to speed up the computation of a local ambient occlusion shading [14]. The top left image in Figure B3.2.9 is rendered using diffuse shading based on the gradient direction. Although high-frequency details of the bones are visible, the relative depth is more difficult to perceive. The top right image shows the result using a local ambient occlusion shading technique. A local model avoids creating fully shadowed parts, which otherwise would be the case in global illumination models. The local ambient occlusion model provides improved depth cues, which is of significant importance. The bottom image
© 2009 by Taylor & Francis Group, LLC
in Figure B3.2.9 shows the combination of diffuse shading and local ambient occlusion. Although further studies are required to explore and evaluate these rendering techniques, it can be seen that these models improve depth perception. In addition, the local ambient occlusion model allows tissue with emissive properties to light up and shade the surrounding voxels. In the example in Figure B3.2.9, the bullet fragments have been made emissive in order to speed up and improve the detection of the bullet and splinters. B3.2.2.4 Virtual Autopsy Workstation Performing a VA provides significant input to the traditional physical autopsy. Many important findings can be discovered and provide valuable evidence and clues in forensic investigations. Our expectation is that VA research will further develop and will focus on addressing validation of forensic pathology procedures and protocols for specific types of VA cases, new imaging modalities such as dual-energy CT and synthetic MRI for improved classification of soft tissue, and novel DVR techniques. Visualization research in the future must include the overall aim of implementing a workstation that includes all approaches needed to perform a VA. Visualization tools to increase quality and efficiency of VA procedures thus need to be developed. From a data management perspective, the DVR component in the VA workstation must be developed to fully capitalize on the multiresolution framework in the PACS platform. Research and development of novel rendering and classification techniques are needed to improve the usability aspects and to specifically address forensic
128
The Virtopsy Approach
FIGURE B3.2.9 Using ambient occlusion and tissue-specific illumination. The 3D structure appears more clear (top right), in contrast to the more flat appearance when only diffuse shading is used (top left). Diffuse shading can, however, provide a more detailed surface structure when such distinct properties are present in the data, combined in bottom image. In the lighting computation the bullet fragments are defined to illuminate their surroundings and emit light in the rendering, which improves the perception of the location of these fragments.
questions. Another important goal is to establish designated protocols for the main forensic case categories. Data analysis research includes the implementation of computer-aided diagnostic tools that, once applied to the postmortem volume data, can help to search for relevant forensic findings, to characterize them, and to deliver general information of the deceased individual such as body height, body weight, sex, major injuries, foreign bodies (e.g., projectiles), and likely causes of death in an automatically generated preliminary written VA protocol.
© 2009 by Taylor & Francis Group, LLC
When these tasks have been addressed, the technology involved in all processes of a VA can be improved to enable advanced decision support and increasing automation of the workflow.
B3.2.3 REFERENCES 1. Johnson, T. R., B. Krauss, M. Sedlmair, et al. 2007. Material differentiation by dual energy CT: initial experience. Eur Radiol 17:1510–17.
3D Visualization of Radiological Data
2. Langheinrich, A. C., A. Michniewicz, D. G. Sedding, et al. 2007. Quantitative x-ray imaging of intraplaque hemorrhage in aortas of ApoE-/-/LDL-/- double knockout mice. Investigative Radiology 42:263–63. 3. Schmidt, T. G., J. Star-Lack, N. R. Bennett, et al. 2006. A prototype table-top inverse-geometry volumetric CT system. Med Phys 33:1867–78. 4. Mazin, S. R., J. Star-Lack, N. R. Bennett, and N. J. Pelc. 2007. Inverse-geometry volumetric CT system with multiple detector arrays for wide field-of-view imaging. Med Phys 34:2133–42. 5. Warntjes, J. B. M., O. Dahlqvist, and P. Lundberg. 2007. Novel method for rapid, simultaneous T1, T2, and proton density quantification. Magn Reson Med 57:528–37. 6. Neeb, H., K. Zilles, and N. J. Shah. 2006. A new method for fast quantitative mapping of absolute water content in vivo. NeuroImage 31:1156–68. 7. Whittall, K. P., A. L. McKay, A. G. Douglas, et al. 1997. In vivo measurements of T2 distributions and water contents in normal human brain. Magn Reson Med 37:34–43. 8. Riederer, S. J., S. A. Suddarth, S. A. Bobman, J. N. Lee, H. Z. Wang, and J. R. McFall. 1984. Automated MR image synthesis: feasibility studies. Radiology 153:203–06.
© 2009 by Taylor & Francis Group, LLC
129
9. Zhu, X. P., C. E. Hutchinson, J. M. Hawnaur, T. F. Cootes, C. J. Taylor, and I. Isherwood. 1994. Magnetic resonance image synthesis using a flexible model. Br J Radiol 67:976–82. 10. Ljung, P. 2006. Efficient methods for direct volume rendering of large datasets. PhD thesis, Linköping University, SE-581 83 Linköping, Sweden. 11. Ljung, P., C. Lundström, and A. Ynnerman. 2006. Multiresolution interblock interpolation in directvolume rendering. In Proceedings Eurographics/IEEE Symposium on Visualization, 259–66. 12. Ljung, P., C. Winskog, A. Perssson, C. Lundström, and A. Ynnerman. 2006. Full body virtual autopsies using a stateof-the-art volume rendering pipeline. IEEE Transactions on Visualization and Computer Graphics (Proceedings Visualization/Information Visualization 2006) 12:869–76. 13. Lundström C., P. Ljung, and A. Ynnerman. 2006. Multidimensional transfer function design using sorted histograms. In Proceedings Eurographics/IEEE-VGTC International Workshop on Volume Graphics. 14. Hernell, F., P. Ljung, and A. Ynnerman. 2007. Efficient ambient and emissive tissue illumination using local occlusion in multiresolution volume rendering. In Eurographics/ IEEE-VGTC International Workshop on Volume Graphics.
B4
Storage of Radiological Data (PACS) Steffen Ross and Michael J. Thali
CONTENTS B4.1 Introduction ....................................................................................................................................................................131 B4.2 Methods of Archiving Cross-Sectional Data..................................................................................................................131 B4.2.1 Archiving on Film .............................................................................................................................................131 B4.2.2 Archiving on CD/DVD......................................................................................................................................131 B4.2.3 Archiving on a PACS.........................................................................................................................................131 B4.2.4 Digital Imaging and Communications in Medicine (DICOM) ........................................................................ 132 B4.3 Equipment ...................................................................................................................................................................... 133 B4.4 How to Purchase a PACS ............................................................................................................................................... 133 References ................................................................................................................................................................................. 133
B4.1 INTRODUCTION
B4.2.1 ARCHIVING ON FILM
In the past decade the use of radiological cross-sectional imaging in forensics has become more common, and a new paradigm has been introduced: the virtual autopsy. At this time only a few institutions in world own a computed tomography (CT) or magnetic resonance (MR) scanner. The cooperation departments of clinical radiology with forensics insitutes is often possible, and this raises the need for a proper archiving method of the externally obtained pictures. With current multislice CT (MSCT) and whole-body MR imaging techniques, it is possible to acquire very large highresolution volumetric datasets. A single MSCT procedure captures about 1 GB of information—and this avalanche of data has to be stored. On the other side, most forensic institutes have made a change from analog to digital photography. Pictures from the crime scene and the autopsy room are saved in dendritic directories, changing a quick search in to digital odyssey. Storage on the Picture Archiving and Communication Systems (PACS) is one possibility toward overcoming these obstacles. All visual information about a single case can be stored as a digital bundle of data (Figure B4.1). With the integration of a database designed for archiving text files, equivalent to the Radiology Information System (RIS), all data belonging to a case can be saved in a single system, which turns a normally time- and nerve-consuming search into a matter of a few mouse clicks.
Printing on film is not suitable for large data volumes. The compromise of going analog by archiving selected pictures is a drawback, because there is no possibility of reevaluating complete datasets for a second opinion, or of retrospective analysis in the case of a scientific study. Storage space is an issue with a growing number of examinations. The retrieval of needed films in endless rows of shelves and the retrieval of a lost film envelope is a well-known and feared problem in clinical radiology.
B4.2 METHODS OF ARCHIVING CROSS-SECTIONAL DATA
PACS is a computer or network system that is used to capture, store, distribute, and then display digital images. Most systems possess an internal backup function, so the danger of losing the archived data is negligible. A PACS consists of four major components (Figure B4.2) [1]:
The three main methods for archiving cross-sectional data are described here.
B4.2.2 ARCHIVING ON CD/DVD Optical storage can also be used as an archive device for storing data. The maximum capacity of a CD is 0.7 GB; a single-sided/single-layer DVD is capable of storing 4.7 GB. The storage capacity of a double-sided/double-layer DVD goes up to 15.9 GB. Besides optical CD/DVD storage, data can be captured on a magneto-optical (MO) disk; its capacity goes up to 9.1 GB. Optical media are sensitive to extremes in temperature, humidity, and direct exposure to sunlight. Although the vendors of medical CDs and DVDs guarantee their products for more than a decade, the stability of the media relies too much on external factors for reliable long-term archiving.
B4.2.3 ARCHIVING ON A PACS
131 © 2009 by Taylor & Francis Group, LLC
132
The Virtopsy Approach
FIGURE B4.1 Diagram of the PACS configuration at the Institute of Forensic Imaging in Bern. Note that the connected digital camera and document scanner can also be an image source for the archiving system.
1. The imaging modalities such as CT and MRI or even a digital camera or document scanner 2. A secured network for the transmission of information 3. Workstations for interpreting and reviewing images 4. Long- and short-term archives for the storage and retrieval of images and reports
B4.2.4 DIGITAL IMAGING AND COMMUNICATIONS IN MEDICINE (DICOM) With the introduction of CT followed by other digital diagnostic imaging modalities in the 1970s and the increasing use of computers in clinical applications, the American College of Radiology (ACR) and the National Electrical Manufacturers
FIGURE B4.2 PACS workstation with the display of pictures from the crime scene and the autopsy room as well as cross-sectional pictures.
© 2009 by Taylor & Francis Group, LLC
Storage of Radiological Data (PACS)
Association (NEMA) recognized the emerging need for a standard method for transferring images and associated information between devices manufactured by various vendors. These devices produce a variety of digital image formats. ACR and NEMA formed a joint committee in 1983 to develop the DICOM standard. It specifies how devices built in conformance with the standards react to commands and data being exchanged. DICOM is a “universal image language,” enabling the communication between different CT/ MR scanners and PACS. It is even possible to archive scanned documents and pictures from a digital camera with the correct patient data by packing them in the DICOM format. The images can all be sent to the same PACS archive, so DICOM is the computer standard that lets a PACS do its work.
B4.3 EQUIPMENT Today it is possible to build an archiving system with common off-the-shelf components. There are different vendors with a variety of PACS solutions on the market. Most of the data to store will come from cross-sectional imaging modalities as CT and MRI. Due to the limited imaging matrix of the two modalities (max 1024 × 1024) there is usually no need to purchase a system with super high-resolution displays needed for conventional radiologic imaging [1]. In contrast to the clinical setup, datasets in forensic imaging are much larger because there are no limitations to the scanned volume by radiation protection issues or the physical fitness of the patient. Thin-sliced whole-body datasets consist of more than 2000 pictures, so the standard PACS workstations used for diagnostic imaging must be adapted to forensic needs with generous updates of computing power, meaning a quite fast processor and the maximum amount of working random access memory (RAM) available. In an ideal system, different functionalities like software modules for multiplanar reconstruction (MPR) and for 3D reconstruction (i.e., shaded surface display [SSD] and volume-rendering technique [VRT]) should already be integrated. So the purchase of a stand-alone reconstruction workstation is not mandatory [2] as long as the PACS workstation fulfills the demands of the user. In addition to “classical” radiologic functionalities, there is the need to archive pictures from digital cameras (taken at the crime scene or in the autopsy room) and microscopic pictures of histopathologic specimens together with the crosssectional data. At the moment there is no out-of-the-box solution for this. A suitable workaround is the installation of a separate program on the workstation that converts the nonradiologic pictures in the DICOM format and attaches them automatically to the previously saved case-related cross-sectional pictures.
© 2009 by Taylor & Francis Group, LLC
133
Besides the commercially available systems, it is possible to build a homemade PACS with free software [3,4] inexpensively. The major drawback of this solution is the lack of guaranteed service in case of a system breakdown. In-house information technology (IT) support that is experienced in medical image networks can fill this gap satisfactorily.
B4.4 HOW TO PURCHASE A PACS A clearly described strategy is fundamental for the painless installation of a PACS. Without proper planning of the system installation, the whole project is doomed to chaos. The following points should be taken into consideration [5–7]: 1. Defining the scope of your PACS: Processing ability of large whole-body datasets is mandatory. 2. Analyzing your workflow requirements so that you can plan workstation deployment: A second, less powerful workstation with minimal reconstruction abilities (axial slices and MPR) is a valuable backup in case of system crash. 3. Ensuring adequate integration in your existing IT structure: The system stands and falls with the support of your IT department. 4. Planning for your image-storage needs: Think big; hard-disk space is inexpensive. 5. Ensuring security of the data transfer: No issue of further discussion in forensic departments. 6. Putting together an effective request for proposal: The more the vendor knows about your needs, the less likely main problems will arise after the installation of the system.
REFERENCES 1. Bennett WF, Vaswani KK, Mendiola JA, and Spigos DG. 2002. PACS monitors: an evolution of radiologist’s viewing techniques. J Digit Imaging 15 (Suppl 1):171–174. 2. Khorasani R. 2005.Buyers beware: Should you purchase a modality or PACS workstation for your practice? What is the difference? J Am Coll Radiol 2:381–382. 3. Rosset A, Spadola L, and Ratib O. 2004. OsiriX: an opensource software for navigating in multidimensional DICOM images. 2004. J Digit Imaging 17:205–216. 4. Rosset C, Rosset A, and Ratib O. 2005. General consumer communication tools for improved image management and communication in medicine. J Digit Imaging 18:270–279. 5. Purchasing a PACS: from planning to procurement. 2005. Health Devices 34:313–324. 6. Branstetter BF. 2007. Basics of imaging informatics: part 1. Radiology 243:656–667. 7. Branstetter BF. 2007. Basics of imaging informatics: part 2. Radiology 244(1):78–84.
B5
The Virtopsy Database: Comparing Radiology and Autopsy Findings Using a Database Emin Aghayev, Lukas Staub, and Michael J. Thali
CONTENTS B5.1 Collaboration with the MEM Research Center ............................................................................................................. 135 B5.2 Virtopsy Database ......................................................................................................................................................... 135 B5.3 Additional Features of the Virtopsy Database .............................................................................................................. 140 B5.4 Privacy Issues ................................................................................................................................................................ 140 B5.5 Unity and Diversity of a Database ................................................................................................................................. 142 B5.6 Summary ....................................................................................................................................................................... 142 References ................................................................................................................................................................................. 142 Since the beginning of the 21st century, forensic medicine has experienced a rapid development of forensic imaging. Recently implemented optical surface documentation and cross-sectional radiological methods are becoming more frequently used as forensic examination tools [1–5]. Comparison of these newer methods with traditional autopsy with regard to specific forensic features has just begun, and a number of advantages have already been reported [1–5]. An important condition for continuous analysis and comparison of new methods with the gold standard is that there be accurate and detailed documentation of the results of both methods regarding different features. For small numbers of cases, a comparison can be easily performed using conventional tables or diagrams, but the growing use of new imaging methods in daily routine makes it impossible to keep track of all findings, features, and examination details. For validation of these methods in forensic medicine, detailed statistical analyses are essential.
B5.1 COLLABORATION WITH THE MEM RESEARCH CENTER The Center for Education and Documentation (MEM-CED) of the Maurice E. Müller Foundation was inaugurated in the 1980s, and its data collection includes information from 1967 onward. Following Müller’s maxim, “No learning without teaching, no learning and teaching without evaluation,” the principle of centralized documentation was developed. Its aims were to serve the needs of training and clinical research through evaluation of case histories, and to facilitate the establishment of a worldwide network of documentation and information systems. As a result, in the 1980s sophisticated documentation and evaluation software intended to assist in the evaluation of outcome studies was designed. This program
and its questionnaires—called International Documentation and Evaluation System (IDES)—were mainly focused on data collection in orthopedic surgery [6–9]. In 2003, MEM-CED was incorporated into the University of Bern, now named Institute for Evaluative Research in Orthopaedic Surgery (IEFO). With its long history and experience dealing primarily with hip- and knee-documentation activities, IEFO has recently extended its core activities to other orthopedic subspecialties as well as to other medical sectors of clinical documentation. Activities are currently centered on a new concept and application in clinical documentation designed and developed by IEFO. Hosted on the memdoc portal [10], this new centralized application offers professionals in the medical sector a convenient and reliable documentation service. Core competencies transcend traditional consultation and setup of clinical trials to a complete one-stop service center using an IEFO-unique application that encompasses many modes of data collection, distribution, and archiving [7–9].
B5.2 VIRTOPSY DATABASE In close collaboration with IEFO, a relational database allowing for Web-based digital documentation of imaging methods and autopsy was implemented (Figure B5.1 and Figure B5.2) [11]. The database is hosted by IEFO and is expected to fulfill many-faceted goals (Figure B5.3), which are becoming or are already topical in forensic medicine. The database was designed according to the concept of centralized documentation provided by IEFO [6–10]. The methodology in centralized data management was built and devised using the latest innovations in medical informatics [9]. 135
© 2009 by Taylor & Francis Group, LLC
136
The Virtopsy Approach
FIGURE B5.1 This figure shows the “New Patient” page of the Virtopsy Module Portal. Users are prompted to fill in general case information such as demographics and dates of incident, death, and inspections. This information is filtered by the module and is not sent through to the central database. See also Figure B5.3.
The outstanding advantage of this methodology is that the end user is not required to purchase, install, or maintain any specialized software or hardware (Figure B5.2). All technical servicing is performed at the central server (Figure B5.2). Accordingly, examination protocols can be easily and quickly distributed to a large user community, while data retrieval and analysis are conducted centrally. Such a system enables an easy setup of local, national, and international registries.
Demographic and case report forms Participants
To participate, end users need only a personal computer with access to the World Wide Web and a user account. The need to purchase hardware arises only when a participating hospital would like to keep full control, and hence responsibility, for all sensitive datasets. In such cases there is a need for a local module server for segregation and anonymization of personal data. Only anonymous data are sent outside of the institution to the central server (Figure B5.2) [11].
Anonymous case report forms Local module #1
Central server Participants
Local module #2
Anonymous case report forms
Demographic and case report forms
FIGURE B5.2 Participants access their own institution’s module that filters demographic data. Only anonymous case report forms are sent to the central server.
© 2009 by Taylor & Francis Group, LLC
The Virtopsy Database: ComparingRadiology and Autopsy Findings Using a Database
137
Goals of the Virtopsy Database
(* . -
') - #$
! "
& +, !
% ! ! !
FIGURE B5.3 Overview of the goals of the Virtopsy Database.
The database consists of questionnaires in which the intake of all forensically relevant findings to all body regions and tissues, to allow for a brief but relatively detailed documentation of forensic cases, was considered (Figure B5.4). A database user is also committed to enter general case information such as the chronology of antemortem radiological examination, incident, death, postmortem radiological examination, and autopsy (Figure B5.4). The questionnaires are broken down into so-called subforms, which represent the amount of data that has to be entered and saved in one documentation step (Figure B5.4 and Figure B5.5). The first three subforms contain questions
on general case information and performed investigations and a steering question in which both the forensic radiologist and forensic pathologist define the affected body parts of a case (Figure B5.4, Figure B5.5, Figure B5.6, and Figure B5.7). With this steering question, the consecutive subforms (i.e., head, neck, thorax, abdomen, spinal cord, soft tissue, and bones) for the forensic pathologist and radiologist are pulled only if they are needed for a specific case (Figure B5.4 and Figure B5.7). The last two subforms of the database contain questions for the forensic radiologist and forensic pathologist, respectively, about manner and cause of death (Figure B5.4).
# #% "&$%
$# (
$&% !$(
) #'
) #'
! #
$
! #
$
%%$$&$
#&$ %
%%$$&$
#&$ %
FIGURE B5.4 The diagram shows the logical design of the Virtopsy Database and its division into subforms. Note the parallel hierarchical placement of the same subforms for forensic pathology and radiology. A difference exists only in the last subform, “manner and cause of death.” This includes 18 additional questions for the forensic pathologist, which are mostly on manner of incident.
© 2009 by Taylor & Francis Group, LLC
138
The Virtopsy Approach
FIGURE B5.5 General case information, which is not sent to the central server.
The workflow-based database validates the quality of the collected data at the point of data entry so that the user cannot submit incomplete, invalid, or inconsistent datasets. Error messages pointing out missing or contradictory answers are displayed, and the data cannot be saved until all of the preprogrammed validation criteria are met (Figure B5.8). This
setup guarantees an accurate, integrative, valid, and competent data sampling and eliminates the need for retrospective data correction. Subforms can be reviewed and corrected as necessary until the entire case has been “submitted” to the central server. After a case has been submitted, it cannot be altered.
FIGURE B5.6 The first five subforms are automatically opened after a new case is initialized (arrows). The second and third subforms (blue arrows) include steering questions for the forensic pathologist and radiologist. See also Figure B5.7.
© 2009 by Taylor & Francis Group, LLC
The Virtopsy Database: ComparingRadiology and Autopsy Findings Using a Database
139
FIGURE B5.7 The steering question for the forensic pathologist triggers consecutive subforms. In this example, all subforms are pulled because all steering questions are clicked on. See also Figure B5.6.
FIGURE B5.8 Example of an online error message that pops up after saving an incomplete subform. According to the message, three questions were incorrectly entered.
© 2009 by Taylor & Francis Group, LLC
140
The enormous human and financial resources needed to key in paper-based data and to correct and complete invalid datasheets are saved using modern Internet and computer technologies. Technically, several information technologies such as Hypertext Preprocessor (PHP), Perl, and Javascript are utilized in the overall architecture of the system, providing a balance between development efficiency and application performance. A Web content management system is utilized to allow parts of the interface to be managed by nontechnical staff.
B5.3 ADDITIONAL FEATURES OF THE VIRTOPSY DATABASE In addition to data submission, the Virtopsy Database provides further useful features for quick treatment of submitted data such as tools for printing, online stats, downloading, and uploading. With the print tool, each form can be printed out in a rough question-and-answer format and added to the case files. A printout of a completely filled out case takes a maximum of five A4 pages and includes a maximum of 157 forensically relevant questions for both the forensic radiologist and pathologist (Figure B5.4). The two investigators—the forensic radiologist and the forensic pathologist—complete the same questions to make future comparisons of radiological and autopsy findings possible. The online stats tool allows users to see descriptive statistics of their own data that are generated by the central SAS
The Virtopsy Approach
engine of IEFO. In addition, these statistics can be compared with the anonymous pool of all virtopsy data within the database (Figure B5.9 and Figure B5.10). With the download tool, users can retrieve their data in a tab-delimited format, which can be imported into all established statistical software packages (Figure B5.11). There is an option to download the whole virtopsy form or just a set of selected items (Figure B5.11). Finally, the picture upload tool includes a possibility of archiving of picture information completive to the case documentation. Radiographic, computed tomographic, photographic, or other images in TIFF or JPEG format can be attached to each case file. The central server is fitted with a Picture Archiving and Communication System (PACS) that allows for automatic reconversion of the images into DICOM format to meet the diagnostic standards of the American College of Radiology, and then displays them via the online interface with various functionalities of picture manipulation (e.g., brightness, contrast, black–white inversion, zooming). However, this feature is not meant to store all images collected for a virtopsy case, as this would require huge amounts of disk space. The number of files per case is currently limited to six images.
B5.4 PRIVACY ISSUES Sending medical data outside the home institution via the World Wide Web presents important challenges with regard to the protection of data privacy. Forensic medical institutions, which deal with different kinds of delinquency, have more problems regarding this issue.
FIGURE B5.9 The figure shows the stats tool, which allows users to analyze their own data online or to compare the data with the pool data (red arrow) by selecting the necessary questions (blue arrows).
© 2009 by Taylor & Francis Group, LLC
The Virtopsy Database: ComparingRadiology and Autopsy Findings Using a Database
141
FIGURE B5.10 As the result of the stats tool, users receive a data report on selected questions in HTML format.
FIGURE B5.11 Using the data-downloading tool, users can retrieve their own data from the database in a tab-delimited format (a). There is an option to download the whole Virtopsy Form or just a set of selected items (blue arrow). The report appears on the same Web page (red arrow), and the data can be then imported into all established statistical software packages, as shown for Microsoft Excel in (b).
© 2009 by Taylor & Francis Group, LLC
142
To address these problems, an additional module server was installed in the Centre of Forensic Imaging and Virtopsy to allow for segregation and encryption of personalized datasets before sending them to the central server. This module server allows Virtopsy Group participants a larger amount of administrative control and accountability over the management of their own data. The module server is a fully self-sufficient interface to the central platform and all its functions [9]. Users can only gain access via the locally hosted module system. While all purely anonymous user and case datasets are submitted to the central database, the module maintains the hosting of all personalized user and case information (Figure B5.2). Among all centralized Web-based registries at the IEFO at the University of Bern, the first forensic registry, the Virtopsy Database, was created under the strictest privacy demands.
B5.5 UNITY AND DIVERSITY OF A DATABASE The Virtopsy Database has been in use since summer 2005 at the Centre of Forensic Imaging and Virtopsy; however, the long-term objective is the establishment of this database for international use. The biggest obstacles in establishing such an interinstitutional collaboration is the heterogeneity of interests and ideas regarding contents and techniques for documentation, and the variety of pathological findings as well as their interpretation. On the other side, there is no doubt that the Internet represents the most ideal and least expensive solution possible to network all users and gather datasets in a central database. Moreover, as already mentioned, other than the module server no costly hardware or software purchases are necessary to run or maintain the installation, since system upgrades and maintenance are conducted only at the central control unit. A Web content management system is utilized to allow parts of the interface to be managed by nontechnical staff members. The generally accepted medical terminology makes the database user friendly and allows for its proper use by residents in radiology and forensic pathology. Of course, the accuracy of entered data depends directly on the professional experience of users. The database questionnaires can presently be entered in English, German, or French; however, translation of the questionnaires into other languages is ongoing. The use of forms and their analysis in different languages is provided because the active fields in the questionnaires refer to a predetermined matrix in the central server application (e.g., in the Interpol protocols of Disaster Victim Identification).
© 2009 by Taylor & Francis Group, LLC
The Virtopsy Approach
B5.6 SUMMARY Studies on virtual autopsy in recent years have formed a new subject in forensic medicine: forensic cross-sectional imaging. After its implementation in forensic medicine and confirmation of its utility and feasibility, the next step should be its validation for routine use. Multiple case studies and close collaboration between forensic institutes, which are essential for validation purpose, can be reached using the Web-based database. The described methodology equipped with the latest innovations in medical informatics serves for a number of important goals of the Virtopsy Database assuring in the same time the strictest protection of data privacy.
REFERENCES 1. http://www.virtopsy.com/publications.htm, 2007. 2. Thali, M. and P. Vock. 2003. Role of and techniques in forensic imaging. In Forensic Medicine: Clinical and Pathological Aspects, ed. J. Payen-James, A. Busuttil, and W. Smock, 731–45. London: Greenwich Medical Media. 3. Thali, M. J., M. Braun, U. Buck, et al. 2005. VIRTOPSY— scientific documentation, reconstruction and animation in forensic: individual and real 3D data based geo-metric approach including optical body/object surface and radiological CT/MRI scanning. J Forensic Sci 50:428–42. 4. Thali, M. J., K. Yen, W. Schweitzer, et al. 2003. Virtopsy, a new imaging horizon in forensic pathology: virtual autopsy by post-mortem multislice computed tomography (MSCT) and magnetic resonance imaging (MRI)—a feasibility study. J Forensic Sci 48:386–403. 5. Aghayev, E., M. Sonnenschein, C. Jackowski, et al. 2006. Fatal hemorrhage in postmortem radiology; measurements of cross-sectional areas of major blood vessels and volumes of aorta and spleen by MSCT and volumes of heart chambers by MRI. AJR 187:209–15. 6. Paterson, D. 1993. The International Documentation and Evaluation System (IDES). Orthopedics 16:11–14. 7. Roder, C., S. Eggli, A. El-Kerdi, et al. 2003. The International Documentation and Evaluation System (IDES)—10-years experience. Internation Orthopaedics 27:259–61. 8. Roder, C., A. El-Kerdi, S. Eggli, and M. Aebi. 2004. A centralized total joint replacement registry using Web-based technologies. J Bone Joint Surg Am 86-A(9):2077–79. 9. Roder, C., A. El-Kerdi, A. Frigg, et al. 2005. The Swiss Orthopaedic Registry. Bull Hosp Jt Dis 63:15–19. 10. http://www.memdoc.org/2007 11. Aghayev, E., L. Staub, R. Dirnhofer, et al. Virtopsy—the concept of a centralized database in forensic medicine for analysis and comparison of radiological and autopsy data. J Forensic Leg Med (submitted).
Part C Forensic Application of Imaging Techniques
© 2009 by Taylor & Francis Group, LLC
C1
Intravital versus Postmortem Imaging Peter Vock
Intravital and postmortem imaging share some common characteristics: They are both nondestructive, easy to document, and allow for the archiving of an examination’s complete information; however, there are also some important differences (Table C1.1). In vivo, the patient is breathing, the circulation and all physiologic processes are going on, and the patient may experience pain, dyspnea, and fear, all potential sources of motion artifacts. Also, side effects of the examination, whether due to radiation exposure or to contrast agents, must be considered and may cause major limitations. In vivo imaging, however, also has significant advantages because it can show image contrast based on functional information in addition to the pure morphology. Most importantly, moving blood is detected by ultrasound, computed tomography (CT), and magnetic resonance imaging (MRI) for angiography, flow quantification, and direction analysis, for perfusion or diffusion measurements, and to differentiate tissues by their specific contrast enhancement. Contrast enhancement means the degree of signal increase or decrease as a function of time after injection of a contrast agent. With most contrast agents, enhancement reflects perfusion, neovascularity, and interstitial diffusion, whether these are iodine-containing benzol derivates in x-ray imaging or gadolinium-containing chelates in MRI. They all have an extracellular distribution volume; from the intravascular space they are distributed into the interstitial space, reaching an equilibrium concentration between the two anatomic compartments and not entering into the cells. Since they are eliminated by glonumerular filtration in the kidneys (and to a minor degree through the biliary system), after a few minutes intravascular concentration starts decreasing, and according to the new concentration gradient, contrast agents diffuse back to the intravascular space. New types of contrast agents are currently being developed with predominantly intravascular distribution for angiography, with biliary excretion for biliary imaging, and with special cellular binding. Molecular imaging will therefore become more and more important to show those specific cell characteristics (i.e., receptors, markers) and to image specific metabolic pathways. Nuclear medicine is currently most advanced in molecular imaging and in general is a functional rather than a morphologic imaging method; to combine both types of information, often functional and morphologic information are now obtained during the same examination (e.g., positron emission tomography [PET]/CT, single photon emission computed tomography [SPECT]/CT). Other types
of functional imaging that are becoming more important in clinical medicine include the demonstration of ventilatory, gravity, and temporal changes where image fusion and 3D postprocessing are important tools to show tiny local functional changes. Postmortem, the lack of circulation and motion is an excellent prerequisite for high quality of morphologic imaging. Measurement time can be prolonged without motion artifacts, and radiation exposure, with its potential for fetal malformation and for cancerogenesis, is not an issue. This allows acquiring signal-to-noise ratios much better than those obtained in vivo, which translates into an excellent image quality. Unfortunately, postmortem imaging also has disadvantages compared with in vivo imaging. First, the delay between the moment of death and imaging is variable, often not exactly known, and the origin of postmortem changes. In other words, differences between a normal examination of a living volunteer and a postmortem study do not necessarily reflect the pathology at the time of death but may also represent ordinary postmortem changes due to gravity (sedimentation), autolysis with a breakdown of barriers, edema and blood extrusion, blood coagulation and degradation, and super infection. For instance, gas in the liver may represent air in a gunshot tract, air embolism as a vital reaction, or gas embolized through the portal circulation from visceral putrefaction. It is obvious that gas is most easily detected by CT. It is also shown by MRI and radiography, and it may severely compromise an ultrasound examination when it lies superficial to the organ of interest. Beyond these changes, postmortem morphologic imaging has a potential similar to in vivo morphologic imaging without contrast agents. The lack of circulation is the reason for the limited use of contrast agents in postmortem imaging. Soft-tissue contrast (excluding gas and calcification) is nearly nonexistent in radiography and is poor in CT; in vivo it can significantly be improved by the easy use of intravascular contrast injection in radiography and CT. Therefore, CT of soft tissues, visceral organs, and vessels is often disappointing in postmortem studies. MRI, due to its high and variable intrinsic contrast, is much better in showing soft tissues and parenchymal and vascular lesions postmortem. But all of these methods, including ultrasound, despite a tomographic analysis may show the distribution of hematoma but not the bleeding site and its quantitative role, which again reflects the purely morphologic rather than functional information. Most nuclear medicine studies are not appropriate postmortem, 145
© 2009 by Taylor & Francis Group, LLC
146
The Virtopsy Approach
TABLE C1.1 Intravital versus Postmortem Imaging Common Advantages of Imaging Methods Nondestructive documentation of the body (original topography) Documentation of complete information at the time of the examination in digital format (no further degradation/putrefaction, no need for fixation of tissue samples) Three-dimensional information using sectional methods Radiography and CT: fast; large volumes; excellent for bone and gas Ultrasound: excellent local/regional information, for superficial soft tissues in case of intact skin and for parenchymal abdominal organs Magnetic resonance imaging: better soft-tissue contrast Potential for image-guided biopsy and treatment causing minimal injury Advantages of Intravital Imaging
Advantages of Postmortem Imaging
Dynamic and functional imaging: motion, blood flow, metabolism (needed for angiography, tissue enhancement, gastrointestinal, and biliary and urinary imaging as well as nuclear medicine techniques) Contrast agents as markers of gastrointestinal tract
Better image quality due to: lack of motion artifacts; higher radiation dose protocols (no side effects by ionizing radiation); no limits by toxicity of and adverse reaction to contrast agents Repeated and whole-body imaging
because the tracer would not reach the organ of its metabolic activity, and physiologic metabolism is replaced by chemical changes due to postmortem degradation. There are some new approaches to overcoming these major limitations. It has been shown that postmortem angiography of excellent quality is feasible, using either a mechanical pump to reestablish circulation in large territories, or local injection of contrast agent after catheterization for one specific vascular area. Since cardiovascular mortality is at the top of the list of causes of death, postmortem angiography of pulmonary embolism and of both coronary artery disease and aortic disease is likely to become an important tool in both pathology and forensic medicine.
© 2009 by Taylor & Francis Group, LLC
The lack of functional information might be overcome noninvasively by MR spectroscopy, a method able to demonstrate relative concentrations of different chemical components in volumes in the range of 1 cm3. Last but not least, as in vivo, image-guided tissue biopsy and fluid aspiration is minimally destructive, is mostly accepted by the relatives of a victim, and provides the material for all types of chemical, bacterial, serologic, and genetic analysis. Despite these differences, intravital and postmortem imaging have many advantages in common, mainly their 3D information, which allows for nondestructive documentation, postprocessing for specific anatomical compartments, and freezing of the exact topographical relation that is usually destroyed at autopsy or surgery.
C2
A Historical Overview of the Literature Lea A. Attias and Richard Dirnhofer
In Forensic Radiology Gil Brogdon points out that examinations of cadavers were begun quite early. It is rewarding, therefore, to precisely study the foundations of the genesis of computed tomography (CT) methods and magnetic resonance (MR) technology. This chapter should make it clear what broad shoulders we are standing on and, at the same time, should serve those who begin with this project and would like to familiarize themselves with the existing literature. On a late Friday evening, November 8, 1895, in his laboratory at the Institute of Physics at Julius Maximilian University in Würzburg, Wilhelm C. Röntgen—as he himself said, “at a time at which no serviceable souls were about”—discovered a previously unknown form of radiation. Röntgen called the radiation “x-rays,” whereby “x” symbolized the unknown. In an 1896 lecture, the famous Anatom Geheimrat von Kölliken suggested that x-rays should be named after their discoverer—namely, Röntgen-rays—and this is what they are called in German. On November 22, 1895, the first x-ray pictures in history were made: With an exposure time of 20 minutes Röntgen photographed the hand of his wife, Bertha Röntgen [1]. On December 28, 1895, Röntgen submitted his manuscript with the title “A New Kind of Rays” to the Physics Medical Society in Würzburg. In this publication Röntgen already spoke about a possible utilization of his discovery in medicine. Röntgen’s discovery had an immediate and powerful impact on forensics because just a few months after the discovery the New York Sun reported that Röntgen had already used this new radiation to photograph broken bones and also projectiles lodged in the body. In January 1896 the Viennese press and the Frankfurter Zeitung also reported concretely and in detail possible clinical applications: among others, that this new technology could be used to display complicated fractures and foreign bodies in a noninvasive way. Only a few months after Röntgen’s announcement, x-ray pictures had already found entry into the courtrooms of North America and England as evidence of projectiles in living and dead persons [2]. This was the birthday of the connection between radiology and forensic medicine—that is, the beginning of clinical-forensic radiology in the imaging of findings on living individuals and forensic-pathologic radiology in examinations of the dead. Also, in Denver, Colorado, within the framework of a civil process this new technology was admitted as being evidential in connection with the question of a treatment error: namely, in the clarification of questions as to whether a fracture was not detected [2]. In Munich,
Angerer employed the x-ray technique to carry out age estimates based on a foot bone x-ray image [2]. Today, forensic age estimation is still based on this, but with hand joint x-ray images according to Greulich and Pyle [3]. In 1896, through intraoral x-ray pictures of teeth, König laid the cornerstone for forensic odontology [2]. One year after the discovery of x-rays, radiological methods were developed by French customs in order to screen packages of suspected smugglers; these machines were the predecessors of the baggage scanners used today in all airports [2]. On December 10, 1901, Röntgen received the first Nobel Prize for physics from the Royal Swedish Academy of Science for his discovery. In the 1920s the first suggestions were made for utilizing x-ray technology to make identifications. Schüller suggested, for example, that a comparison of sinus cavities could be used for identification purposes [4]. Subsequently, a rapid broadening of the forensic-radiological examination spectrum took place. This mainly pertained to the forensic interpretation of skeletal injuries, child abuse, torture, identification in the event of mass catastrophes, and forensic odontology, as well as proving the existence of foreign objects [3–34]. From the field of forensic ballistics it is worth mentioning that Lorenz, in the 1948 German Journal of Legal Medicine, published an overview concerning projectile trajectories in conventional x-ray pictures [17]. Subsequently, Thali and Dirnhofer point to the technical quantum leap, Hounsfield’s so-called cross-section scanning CT procedure that led to numerous postmortem studies [1]. From the field of forensic ballistics, the Berlin paper by Wüllenberger, Schneider, and Grumme concerning skull injuries [35] is worth highlighting. This represents the beginning of forensic-radiological cross-sectional imaging on living persons. In their work, the authors point out that CT examinations are superior to conventional procedures, primarily in view of the possibility of reconstruction of bullet trajectories within the body. In a later Berlin study that used a CT image, Schneider, Woweries, and Grumme (1979) investigated the shaking trauma of an infant [36]. Soon afterward, in 1982, Schmidt and Kallieris from Heidelberg published the first overview of postmortem radiological examining techniques in their paper “Use of Radiographics in the Forensic Autopsy” [37]. In the same year Bratzke, Schneider, and Dietz published a review of x-ray examinations in an inquest autopsy wherein they pointed out that a postmortem CT could possibly play a role in situations where sectioning is forbidden [38]. The idea for comprehensive postmortem examinations of gunshot 147
© 2009 by Taylor & Francis Group, LLC
148
The Virtopsy Approach
wounds using CT was proposed for the first time by Clasen and Torack in their 1982 paper “Computerized Tomography and Neuropathology” [39]. In 1983 Schumacher, Oehmichen, König, Einighammer, and Bien published investigations around the theme of CT studies on wound ballistics of cranial gunshot injuries, in which the results of intravital CT examinations in cranial gunshot injuries were compared with the corresponding neuropathologic sections [40,41]. They came to the conclusion that imaging methods such as CT and ultrasound are valuable supplements to neuropathological examining in forensic questions and should be applied before neuropathologic workup of the brain. Moreover, their work showed that, with the CT technology of that early stage of its development, the entries and exits as well as trajectories of gunshot wounds could also be displayed. In 1989 Riepert systematically published numerous papers concerning forensic radiology that are viewed today as basic forensic radiological investigations. His studies primarily pertain to the area of identification and image processing, as well as age determinations [42–62]. Starting in the mid 1990s, at the Institute for Forensic Medicine of the University of Heidelberg, K. M. Stein occupied himself intensively with CT diagnostics and their possible forensic applications [63–71]. The first impressive 3D visualizations of head injuries as well as the first examinations for determining gunshot residues with CT arose there. In 1989 Amberg published a paper about making identifications that he carried out using magnetic resonance imaging (MRI) [72]. In addition, after the mid 1990s, Stein presented initial examination results regarding the utilization of MRI in strangulation cases [73]. In 1998, MRI data was used in Heidelberg for planning of the subsequent autopsies of twins who had died [74]. Also in 1998 Karger, from the Institute for Forensic Medicine of the University of Münster, published MRI organ examinations of the brain in connection with experimental gunshot injuries [75] and, shortly afterward, 3D CT visualizations of cranial gunshot injuries of living persons [76]. In 2000, Hess and Harms, from the Technical University of Munich, published a groundbreaking review of MR and the evaluation of gunshot wounds [77]. From 1974 to 1999 numerous forensic-radiological studies were made [73–95]. In the 1990s all of these preliminary achievements and perspectives influenced the Institute of Forensic Medicine at the University of Bern in its decision to make forensic imaging the main focus of its research efforts.
REFERENCES 1. Roentgen, W. C. 1895. Ueber eine neue Art von Strahlen. Sitzungsberichte der physikalisch-medizinischen Gesellschaft, p. 132. [About a new kind of rays. Proceedings of the Physical-Medical Society.] 2. Brogdon, B.G. 1998. Forensic Radiology. Boca Raton, FL: CRC Press.
© 2009 by Taylor & Francis Group, LLC
3. Greulich, W. and S. Pyle. 1959. Radiographic Atlas of Skeletal Development of the Hand and Wrist, 2d ed. Palo Alto, CA: Stanford University Press. 4. Schüller, A. 1921. Das Röntgenogramm der Stirnhöhle. Ein Hilfsmittel für die Identitätsbestimmung von Schädeln. Mschr Ohrenheilk 55:1617–20. 5. Ottolenghi, S. 1899. Ein neues Todeszeichen and der Einfluss der Respiration and der Verwesung auf die Radiographie der Lungen. Vschr Ger Med Oeff San 17:282–88. 6. Troeger. 1903. Über Röntgenstrahlen in gerichtlich-medizinischer Beziehung. Friedreichs Blätter Ger Med San Pol 4:241–79. 7. Kenyers, B. 1907. Mitteilungen zu gerichtsärztlichen Beurteilungen von Röntgenbildern. Vschr Ger Med 34:89–92. 8. Schwarz, L. 1909. Die Bedeutung der Röntgenstrahlen für die gerichtliche Medizin. Fortschr Röntgenstr 13: 191–231. 9. Fränkel, P. and H. Marx. 1913. Aufklärung der Todesursache durch Röntgenstrahlen. Arch Kriminol 54:103–10. 10. Hildebrand. 1914. Das Röntgenverfahren in der gerichtlichen Medizin. In Gerichtsärztliche and polizeiärztliche Technik, ed. T. Lochte. Wiesbaden: J.F. Bergmann. 11. Bucky, G. 1922. Kriminalistische Fragestellungen durch Röntgenstrahlen. Aerztl Sachv Z 28:166–70. 12. Richter, H. 1926. Ein Beitrag zur Bedeutung des Röntgenverfahrens in Kriminalfällen. Dtsch Z ges gerichtl Med 7:626–33. 13. Izkovitsch, I. 1930. Röntgenologie im Dienste des Gerichtes. Fortschr Röntgenstr 42:664–66. 14. Izkovitsch, I. 1936. Röntgenologische Altersbestimmung für Gerichtszwecke. Fortschr Röntgenstr 54:249–50. 15. Schuller, A. 1943. A note on the identification of skulls by x-ray pictures of the frontal sinuses. Medical Journal of Australia 1:555. 16. Berndt, H. 1947. Entwicklung einer röntgenologischen Altersbestimmung am proximalen Humerusende. Z Ges Inn Med 2:122–28. 17. Lorenz, R. 1948. Der Schusskanal im Röntgenbilde. Dtsch Z ges gerichtl Med 39:435–48. 18. Holczabek, W. 1955. Ein Beitrag zur Identifikation durch vergleichende Röntgenuntersuchung. Beitr Gerichtl Med 20:35–36. 19. Kellner, H. 1957. Untersuchungen zur röntgenologischen Altersbestimmung am proximalen Humerusende bei Erwachsenen. Universität Bonn. 20. Leopold, D. and G. von Jagow. 1961. Das Röntgenbild des Kehlkopfes—eine Möglichkeit zu Altersbestimmungen. Beitr Gerichtl Med 21:181–90. 21. Neiss, A. 1961. On osteological details visible in roentgen pictures, but little known in anatomy. Anat Anz 110:102–15. 22. Neiss, A. 1961. On little-known skeletal variations. Fortschr Geb Roentgenstr Nuklearmed 94:227–32. 23. Neiss, A. 1961. Die Aufgaben der Röntgenologie nach Flugzeugunglücken. Z Kriminalistik 15:343–44. 24. Neiss, A. 1962. The problems of roentgen anthropology. Fortschr Geb Roentgenstr Nuklearmed 97:57–62. 25. Neiss, A. 1962. Röntgen-Identifikation. Wehrmed Mitteilungen 4:49–52. 26. Neiss, A. 1964. Röntgenidentifikation durch Bildvergleiche. Z Ges Ger Med 55:135–36. 27. Neiss, A. 1966. Poachers convicted with aid of x-rays. Fortschr Geb Roentgenstr Nuklearmed 104:428.
A Historical Overview of the Literature
28. Neiss, A. 1967. On the significance of x-ray picture files as documentation of phylogeny. Anat Anz 121:396–400. 29. Neiss, A. 1968. Röntgenidentifikation. Stuttgart: Thieme. 30. Krause, D. 1968. Zur Identifikation unbekannter Leichen durch Röntgenbildvergleiche. Beitr Gerichtl Med 24:36–41. 31. Leopold, D. 1968. Identifikation durch Schädeluntersuchung unter besonderer Berücksichtigung der Superprojektion. Universität Leipzig. 32. Neiss, A. 1968. Bei Verdacht auf Tötung durch Schuss sollte geröntgt werden. Fortschr Röntgenstr 109:668–69. 33. Neiss, A. 1969. Sollen Fundleichen geröntgt werden? Z Kriminalistik S 414–17. 34. Fischer, H., H. Masel, and J. Steinberg. 1970. Postmortem radiography with the aid of a field x-ray machine. Fortschr Geb Roentgenstr Nuklearmed 113:535–37. 35. Wullenweber, R., V. Schneider, and T. Grumme. 1977. A computer-tomographical examination of cranial bullet wounds (author’s transl). Z Rechtsmed 80:227–46. 36. Schneider, V., J. Woweries, and T. Grumme. 1979. Trauma inflicted on a baby by shaking (author’s transl). MMW Munch Med Wochenschr 121:171–76. 37. Schmidt, G. and D. Kallieris. 1982. Use of radiographs in the forensic autopsy. Forensic Sci Int 19:263–70. 38. Bratzke, H., V. Schneider, and W. Dietz. 1982. Radiographic investigation during medico-legal autopsies (author’s transl). ROFO Fortschr Geb Roentgenstr Nuklearmed 136:463–72. 39. Clasen, R. A., and R. M. Torack. 1982. Computerized tomography and neuropathologists: two viewpoints. J Neuropathol Exp Neurol 41:387–88. 40. Schumacher, M., M. Oehmichen, H. G. KÖnig, and H. Einighammer. 1983. Intravital and Postmortal CT examinations in cerebral gunshot injuries. ROFO Fortschr Geb Roentgenstr Nuklearmed 139:58–62. 41. Schumacher, M., M. Oehmichen, H. G. KÖnig, H. Einighammer, and S. Bien. 1985. Computer tomographic studies on wound ballistics of cranial gunshot injuries. Beitr Gerichtl Med 43:95–101. 42. Riepert, T. and C. Rittner. 1989. Roentgen identification of unknown cadavers and living persons. Beitr Gerichtl Med 47:207–14. 43. Riepert, T. and C. Rittner. 1989. Roentgen identification of unknown cadavers with advanced postmortem changes. Z Rechtsmed 102:207–16. 44. Riepert, T., G. Endres, M. Knapp, and C. Rittner. 1991. Röntgenidentifizierung durch elektronische Bildmischung. Rechtsmedizin 1:51–7. 45. Riepert, T., R. Mattern, C. Rittner, and H. Schild. 1992. Postmortem Röntgendiagnostik des Thorax im rechtsmedizinischen Obduktiongut unter besonderer Berücksichtigung von Fehlerquellen. Rechtsmedizin 3:1–5. 46. Lasczkowski, G., T. Riepert, and C. Rittner. 1992. Site of discovery in the bath tub. Evaluation of a fatality after four years using postmortem roentgen diagnosis. Arch Kriminol 189:25–32. 47. Riepert, T., D. Ulmcke, U. Jendrysiak, and C. Rittner. 1995. Computer-assisted simulation of conventional roentgenograms from three-dimensional computed tomography (CT) data—an aid in the identification of unknown corpses (FoXSIS). Forensic Sci Int 71:199–204. 48. Riepert, T., C. Rittner, D. Ulmcke, S. Ogbuihi, and F. Schweden. 1995. Identification of an unknown corpse by means of computed tomography (CT) of the lumbar spine. J Forensic Sci 40:126–27.
© 2009 by Taylor & Francis Group, LLC
149
49. Riepert, T., F. Schweden, H. Schild, and C. Rittner. 1995. The identification of unknown corpses by x-ray comparison. ROFO Fortschr Geb Roentgenstr Neuen Bildgeb Verfahr 162:241–45. 50. Riepert, T. 1995. Sekandäre Röntgenidentifizierung unter besonderer Berücksichtigung digitaler Bildverarbeitung and forensischer Relevanz. Universität Mainz. 51. Riepert, T., T. Dretscher, H. Schild, B. Nafe, and R. Mattern. 1995. Zur Röntgenmorphologie der Ferse unter besonderer Berücksichtigung der Häufigkeit von Varianten im Hinblick auf eine Individualisierung. Rechtsmedizin 5:142–46. 52. Riepert, T., T. Dretscher, H. Schild, B. Nafe, and R. Mattern. 1996. Estimation of sex on the basis of radiographs of the calcaneus. Forensic Sci Int 77:133–40. 53. Riepert, T., C. Neumann, F. Schweden, and R. Urban. 1996. Identification of unknown cadavers in forensic medicine practice. Arch Kriminol 198:23–30. 54. Kreitner, K. F., F. Schweden, H. H. Schild, T. Riepert, and B. Nafe. 1997. Computerized tomography of the epiphyseal union of the medical clavicle: an auxiliary method of age determination during adolescence and the 3rd decade of life? ROFO Fortschr Geb Roentgenstr Neuen Bildgeb Verfahr 166:481–86. 55. Riepert, T., T. Dretscher, H. Schild, B. Nafe, and R. Mattern. 1998. Zur säkulären Akzeleration des Calcaneus—Analyse auf der Grundlage von Röntgenaufnahmen der Ferse. HOMO 49:13–20. 56. Riepert, T., A. Schultes, C. Berchtenbreiter, K. Lackner, and M. Staak. 2000. Pathomorphologische Befunderhebung vs computertomographische Befunderhebung—ein prospektiver Methodenvergleich. Rechtsmedizin 10:47. 57. Schultes, A., A. Schuff, H. Grass, T. Riepert, and K. Lackner. 2001. Hilfe durch postmortem CT bei der Klärung der Todesursache: Darstellung anhand zweier Obduktionsfälle. Rechtsmedizin 11:158–59. 58. Riepert, T., A. Schultes, H. Grass, et al. 2001. Autopsie und postmortale Computertomographie—ein prospektiver Vergleich. Rechtsmedizin 11:160. 59. Riepert, T., D. Ulmcke, F. Schweden, and B. Nafe. 2001. Identification of unknown dead bodies by x-ray image comparison of the skull using the x-ray simulation program FoXSIS. Forensic Sci Int 117:89–98. 60. Rothschild, M., B. Krug, and T. Riepert. 2001. Postmortale Röntgendiagnostik in der Rechtsmedizin. Rechtsmedizin 11:230–43. 61. Riepert, T., R. M. Flores, and R. Urban. 2001. Postmortem roentgen diagnosis of the skull. Arch Kriminol 208: 65–71. 62. Grass, H., T. Riepert, J. Falk, M. Löhr, and B. Krug. 2003. Kraniale Computertomographie versus Sektionsbefund. Rechtsmedizin 13:96–99. 63. Stein, K., L. Bahner, and R. Mattern. 1997. 3-Dimensionale Schusskanalrekonstruktion mittels Computertomographie. Paper presented at 27. Treffen der Oberrheiner Rechtsmediziner, Basel. 64. Stein, K., M. Bahner, J. Merkel, S. Ain, and G. Heilmann. 1997. Computertomographische Einschussmorphologie von Bleigeschossen, eine experimentelle Studie. Paper presented at 76. Jahrestagung der Deutschen Gesellschaft für Rechtsmedizin, Jena. 65. Stein, K., L. Bahner, and R. Mattern. 1998. 3-Dimensionale Schusskanalrekonstruktion mittels Computertomographie. Fortschr Röntgenstr 168:216. 66. Stein, K., L. Bahner, J. Merkel, S. Ain, and G. Heilmann. 1998. Computertomographische Einschussmorphologie von Bleigeschossen—eine experimentelle Studie. Fortschr Röntgenstr 168:216.
150
67. Stein, K., T. Hertle, S. Ain, L. Bahner, and R. Mattern. 1999. Computertomographische Differenzierung von Geschossarten—eine experimentelle Studie. Rechtsmedizin 9 (Suppl 1). 68. Stein, K. M., M. L. Bahner, J. Merkel, S. Ain, and R. Mattern. 2000. Detection of gunshot residues in routine CTs. Int J Legal Med 114:15–18. 69. Stein, K., K. Ruf, D. Mtejic, and R. Mattern. 2003. Methoden zur Darstellung subduraler Blutungsquellen mit der postmortemn Computertomographie. Rechtsmedizin 194:228–29. 70. Stein, K. 2003. Computed tomography in forensic pathology, what is a normal finding in cadaver. Forensic Sci Int 136:270. 71. Schnatterberg, P., K. Stein, L. Bahner, C. Rehm, W. Reith, and G. van Kaick. 1996. Lungenkontusionen: eine postmortem radiologisch-histopathologische Studie. Fortschr Röntgenstr 164:154. 72. Amberg, R., B. Foster, and Fürmaier. 1989. Identification by MRI. Z Rechtsmed 102:185–89. 73. Stein, K., S. Delorme, H. Schlemmer, R. Mattern, and E. Miltner. 1997. Ultrasound and MRI examination in forensic medicine. A method to detect and quantify traces of recent strangling. Paper presented at 17th Congress of the International Academy of Legal Medicine, Dublin. 74. Sergi, C., A. Dorfler, F. Albrecht, et al. 1998. Utilization of magnetic resonance imaging in autopsy planning with specimen preservation for thoraco-omphalopagus symmetricus conjoined twins. Teratology 58:71–75. 75. Karger, B., Z. Puskas, B. Ruwald, K. Teige, and G. Schuirer. 1998. Morphological findings in the brain after experimental gunshots using radiology, pathology and histology. Int J Legal Med 111:314–19. 76. Karger, B., W. Heindle, G. Fechner, and B. Brinkmann. 2001. Proof a gunshot wound and its delayed effects 54 years post injury. Int J Legal Med 115:173–75. 77. Hess, U. and J. Harms. 2000. Die MRT zur Beurteilung von Schussverletzungen. Rechtsmedizin 10:90–95. 78. Bohnert, M., A. Schmialch, M. Faller-Marquardt, C. BuitragoTellez, and S. Pollak. 1998. Umwandlung des Gehirns and der Gesichtsweichteile in Leichenlipid–morphologische and radiologische Befunde. Rechtsmedizin 8:135–40. 79. Buitrago-Tellez, C., W. Wenz, G. Freiedrich. 1992. Digitale Röntgenbildbearbeitung als Hilfsmittel in der Rechtsmedizin. Radiologe 32:87–89. 80. Eckert, W. G. and N. Garland. 1984. The history of the forensic applications in radiology. Am J Forensic Med Pathol 5:53–56. 81. Jachau, K., W. Döhring, and D. Krause. 1999. Identification eines skelettierten Schädels durch CT-Bildvergleich und Superimposition. Paper presented at 8. Frühjahrstagung der DGRM–Region Nord.
© 2009 by Taylor & Francis Group, LLC
The Virtopsy Approach
82. Madea, H., N. Tanaka, and H. Ohshima. 1989. Roentgenographic identification of unknown human remains. Two case reports. Act Crim Jap 55:164–70. 83. Markert, K. and I. Wirth. 1982. Elektronische Bildmischung—eine Möglichkeit zur Optimierung des Vergleichs von prä- and postmortemn Röntgenbildern. Kriminalistik forens Wiss 47:11–16. 84. Metter, D., H. Stute, and J. Rauschke. 1989. Forensic medicine roentgen studies in gunshot wounds of the skull. Beitr Gerichtl Med 47:473–77. 85. Penning, R., W. Eisenmenger, and W. Spann. 1995. Röntgenstrahlen in der Rechtsmedizin. In Forschung mit Röntgenstrahlen—Bilanz eines Jahrhunderts (1895–1995), ed. F. H. W. Heuck and E. Macherauch, 257–71. Berlin: Springer. 86. Penning, R., W. Eisenmenger, and W. Spann. 1998. Zum Einsatz radiologischer Methoden in der Rechtsmedizin. In Klinische Radiologie—Diagnostik mit bildgebenden Verfahren, ed. F. H. W. Heuck and E. Macherauch, 533–45. Berlin: Springer. 87. Pollak, S. and A. Lindermann. 1990. Verletzungsbilder und Röntgenbefunde nach Schüssen mit selten verwendeter Flintenmunition. Beitr Gerichtl Med 48:507–18. 88. Schmidt, G. and D. Kallieris. 1980. Rechtsmedizinische Röntgenuntersuchungen. Zentralbl ges Rechtsmed 20:42. 89. Schmidt, G. 1986. Radiologischer Nachweis der Luftembolie. In Medizin and Recht. Festschrift für Wolfgang Spann, ed. W. Eisenmenger, E. Liebhardt, and M. Schuck. Berlin: Springer. 90. Schmidt, G., D. Kallieris, and P. Stosch. 1990. Roentgenologic heart volume determination for the diagnosis of fatal heart failure. Z Rechtsmed 103:155–62. 91. Schneider, V. 1978. Aneurysmablutung mit Einbruch in das Ventrikelsystem—computertomographische Befunde and ihre pathologisch-anatomischen Korrelate. Zacchia 53:176. 92. Vogel, H. 1997. Gewalt im Röntgenbild: Befunde im Krieg, Folter and Verbrechen. Landsberg/Lech: Ecomed. 93. Zink, P., I. Zink, and G. Reinhardt. 1986. Roentgen image of the hand as a principle for age determination in adolescents. Arch Kriminol 178:15–24. 94. Freislederer, A., W. Bautz, and V. Schmidt. 1988. Body packing: the value of modern imaging procedures in the detection of intracorporeal transport media. Arch Kriminol 182:143–53. 95. Freislederer, A., W. Sauer, M. Graw, and V. Schmidt. 1998. Body-Packing: Nachweis inkorporierter Drogenpäckchen mittels Ultraschalltechnik. Beitr Gerichtl Med 47:187–91.
C3
External Body Documentation Silvio Näther, Ursula Buck, and Michael J. Thali
CONTENTS C3.1 Photogrammetric Analysis of Injuries........................................................................................................................... 151 C3.2 Generation of 3D Models from the Surface Scanning Data .......................................................................................... 151 C3.2.1 Alignment ......................................................................................................................................................... 151 C3.2.2 Polygonization .................................................................................................................................................. 151 C3.2.3 Mesh Smoothing and Thinning ........................................................................................................................ 153 C3.3 Visualization of the 3D Models in True Color .............................................................................................................. 153 References ................................................................................................................................................................................. 156
C3.1 PHOTOGRAMMETRIC ANALYSIS OF INJURIES Various photogrammetric software is used for multiple image evaluation; for example RolleiMetric CDW [1], Elcovision [2], TRITOP [3], iWitness [4], etc. The basic principles of photogrammetric evaluation are shown based on the TRITOP software. From the photogrammetric images, the computer software calculates the image coordinates, the camera positions, and the 3D coordinates of the measuring points. For this calculation, the software needs to know two or three reference distances, the type of camera used as well as its approximate focal length. The reference targets are measured automatically, by the use of coded reference targets, or manually, measuring every reference point in three or more images. The results give the calculated image positions and the 3D coordinates of the measuring points. The three analyzing functions are to measure the injury lines, the important points, and the distances, automatically or manually (Figure C3.3.1). The calculated images can also be projected onto the object surface to get a true color 3D model. The results can be exported into standard formats for further analysis, for example comparison with the injurycausing instrument, or they can be loaded directly into the ATOS scanning software.
C3.2 GENERATION OF 3D MODELS FROM THE SURFACE SCANNING DATA The result of the optical surface scanning of a measured object is a point cloud of millions of points—each scan delivering up to 4 million—with three-dimensional coordinates that represent the scanned object. To work with the data (e.g.,
editing or measuring dimensions) they have to be processed. The steps for the generation process of the 3D models, including the alignment, polygonization, mesh smoothing, and thinning, are described here.
C3.2.1 ALIGNMENT The individual measurements are transformed into a common coordinate system by the reference points applied to the object. By means of fine alignment, deviations between the individual measurements are reduced. The alignment runs automatically, but the operator has the option of adjusting parameters that influence the processing. Even after fine alignment, randomly distributed deviations remain in the overlapping areas of the individual measurements that exist due to the measuring noise of the points. A color view and a scale represent the deviations (Figure C3.3.2). When the deviations are acceptable, the point clouds can be polygonized.
C3.2.2 POLYGONIZATION Polygonization means the conversion of the measuring point cloud into a mesh of nonoverlapping triangles. The mesh has different densities depending on the curvature of the object. The operator can influence the density of the mesh by adjusting the so-called polygonization raster. To specify the density of the polygon mesh one can adjust the ratio of points in strong curved and plane areas that are included in the calculation. For small parts with tiny details, a ratio of 1:1 will be selected; for larger parts with small details ratios of 1:2 or 1:4 are selected; and for larger parts a ratio of 2:8 will be selected. With the raster setting of, for example, 1:4, every point is used in high curvature areas, every second point is used in medium
151 © 2009 by Taylor & Francis Group, LLC
152
The Virtopsy Approach
FIGURE C3.3.1 Photogrammetric analysis of injuries: Measurement of 3D points and lines in two or more images. Epipolar lines (green line) help to locate the identical point in the second or the following image.
FIGURE C3.3.2 With the alignment of all single measurements, remaining small gaps are closed. The mesh deviation is displayed in the 3D color view, with values in the dialog box (red ellipse). It represents the remaining noise of all meshes.
curvature areas, and every fourth point us used in flat areas. This adjustment is important to find the balance between possible low data volume and adequate resolution. If necessary, the gaps in the data that are created by the reference targets
can be filled automatically. During the polygonization process the individual measurements are finely adjusted to each other and are recalculated with the highest point resolution. The overlapping areas are deleted and merged to a polygon mesh (Figure C3.3.3A,B; Figure C3.3.4A).
A
C
B
D
E
FIGURE C3.3.3 Representation of the polygon mesh structure of the 3D model: (A) Polygonized mesh of the backside of a body with a raster setting of 1:1. (B) Polygon mesh in detail (red rectangle in 3a). (C) Display of the triangles of the detailed mesh. (D) After smoothing the mesh to eliminate the measurement noise. (E) After thinning the mesh without loss of important details (red arrow signals a hair lying on the body; the yellow arrow points to a strong thinned area of the cut and filled reference marker).
© 2009 by Taylor & Francis Group, LLC
External Body Documentation
A
153
B
C
FIGURE C3.3.4 (A) Polygonized mesh of a small skin injury with a raster setting of 1:1 (length of 5 mm). (B) After smoothing the mesh to eliminate the measurement noise. (C) After thinning the mesh without loss of important details.
C3.2.3 MESH SMOOTHING AND THINNING Smoothing eliminates existing measuring noise. During smoothing the points are shifted so that they integrate better on the surface. The operator can influence the smoothing by adjusting several parameters. A smoothing intensity that eliminates the point noise but keeps small object details that are important for the analysis should be chosen (Figure C3.3.3C,D). For certain applications (e.g., visualization and animation) the amount of polygon mesh data (3D model) should be reduced by thinning. The thinning takes into consideration the curvature of the object. Strong curvature areas are thinned less than flat areas. This ensures that important details are kept (Figure C3.3.3E, Figure C3.3.4B,C). The operator can influence the thinning by several parameters, such as maximum edge length of the triangles. For the comparison of an inflicting tool to a patterned injury, many details are needed. However, the mesh should not be thinned too much in this area, nor should there be too much smoothing; both procedures can negatively influence the desired data. For other purposes (e.g., animation for accident reconstruction) a low amount of data is more important than high detail.
C3.3 VISUALIZATION OF THE 3D MODELS IN TRUE COLOR For the forensic documentation of external injuries, the color of the injury and skin as well as the shape plays an important role. The pattern of an injury is often presented by color transitions. Therefore, a true-color 3D model of an injury offers valuable clues to the injury-inflicting instrument or the
© 2009 by Taylor & Francis Group, LLC
kind of impact. On this account, the high-quality, colored 3D data are generated through projecting color information from images onto digitized surface data. There are two methods used, which are both described herein. The first method is to use the captured photogrammetric images calculated in the TRITOP software. These images, which receive the color information for the surfaces, are taken in addition to the normal TRITOP images after removing the coded crosses that covered the body surface. The resulting polygon mesh from the scanning process is imported into the TRITOP photogrammetry system. On this imported 3D model, one or more images can be automatically projected onto the surface. In this process, every triangulation point gets a color value from one or a mix of more images. The resulting data file includes excellent shape and color definition, as can be seen in Figure C3.3.5, Figure C3.3.6, and Figure C3.3.7. The second method is to project well-illuminated photographs, which are taken of the object with reference targets, over the 3D model with the animation software 3D studio max [5]. By means of the reference targets, measured in the 3D model as well as in the photographs, the camera position and the focal lens are calculated. The photographs are defined as a material and are projected over the 3D surface model using the required camera information (Figure C3.3.8). The result is high-resolution color information not dependent on the optimizing degree of the model. This affords the possibility of optimizing the surface data for a faster working process, and of obtaining a high resolution from the texture material. This second method is best used for real data-based animation with complex datasets and for the documentation of living persons when no photogrammetry is implemented (Figure C3.3.9).
154
The Virtopsy Approach
FIGURE C3.3.5 Gray-shaded polygon mesh of the front side of the body resulted from the optical surface scan data.
FIGURE C3.3.6 3D surface models of the whole body of the accident victim with all relevant patterned injuries true to color.
© 2009 by Taylor & Francis Group, LLC
External Body Documentation
155
FIGURE C3.3.7 The high-resolution, true-to-color surface model displays a minute skin abrasion (red arrow; see also Figure C3.3.4), the structure of the skin, and a hair lying on the body (yellow arrows). A
B
FIGURE C3.3.8 3D model of a bite mark on the upper arm of a living person: (A) High-resolution surface data resulted from digitizing with the ATOS III. (B) The true-to-color 3D surface data of the bite mark, generated by projecting a digital photograph onto the gray-shaded 3D model.
FIGURE C3.3.9 Different views of a true-to-color 3D model of the face of a living person.
© 2009 by Taylor & Francis Group, LLC
156
The Virtopsy Approach
REFERENCES 1. http://www.rolleimetric.com, Braunschweig, Germany, 2007. 2. http://www.elcovision.com, St. Margrethen, Switzerland, 2007.
© 2009 by Taylor & Francis Group, LLC
3. http://www.gom.com, Braunschweig, Germany, 2007. 4. http://www.iwitnessphoto.com, Melbourne, Australia, 2007. 5. http://www.autodesk.com/3dsmax, USA, 2007.
C4
Internal Body Documentation
CONTENTS C4.1 Conventional Radiology ................................................................................................................................................ 157 C4.1.1 References......................................................................................................................................................... 157 C4.2 Forensic Ultrasonography .............................................................................................................................................. 157 C4.2.1 References......................................................................................................................................................... 157 C.4.3 Computed Tomography (CT) and Magnetic Resonance Imaging (MRI) ..................................................................... 157 C.4.4 Magnetic Resonance Imaging (MRI) ............................................................................................................................ 157
C4.1 CONVENTIONAL RADIOLOGY
C4.2 FORENSIC ULTRASONOGRAPHY
Lars Oesterhelweg and Michael J. Thali
Lars Oesterhelweg and Michael J. Thali
A detailed presentation of the conventional radiological methods in the use of forensic radiology will not be given in this book. We recommend Gil Brogdon’s Forensic Radiology [1], which gives a detailed and complete overview of the topic. Furthermore, the Radiological Atlas of Abuse, Torture and Inflicted Trauma by Brogdon, Vogel, and McDowell should be mentioned [2]. In the forensic field of child abuse, an excellent textbook by Paul Kleinmann is available [3]. Another controversial issue in the field of application of conventional radiology in forensic science is the estimation of the age of living people undergoing criminal proceedings [4,5]. Therefore, in the relevant age groups, radiographs of the hand and the sternoclavicular joint as well as a panorthotomograph are recommended.
The Virtopsy Project has not yet undertaken any studies in postmortem ultrasonography. A Spanish group that created the term echopsy [1] and a Japanese–German group under the lead of Seisaku Uchigasaki have performed studies in this field of imaging application [2,3].
C4.1.1 REFERENCES 1. Brogdon, B. G. 1998. Forensic radiology. Boca Raton, FL: CRC Press. 2. Brogdon, B. G., H. Vogel, and J. D. McDowell. 2003. A radiological atlas of abuse, torture, terrorism, and inflicted trauma. Boca Raton: CRC Press. 3. Kleinmann, P. K. 1998. Diagnostic imaging of child abuse, 2d ed. Baltimore: Mosby. 4. Schmeling, A., G. Geserick, W. Reisinger, and A. Olze. 2007. Age estimation. Forensic Sci Int 165:178–81. 5. Schmeling, A., A. Olze, W. Reisinger, and G. Geserick. 2001. Age estimation of living people undergoing criminal proceedings. Lancet 358:89–90.
C4.2.1 REFERENCES 1. Fariña, J., C. Millana, M. J. Fdez-Aceñero, et al. 2002. Ultrasonographic autopsy (echopsy): a new autopsy technique. Virchows Arch 440:635–39. 2. Uchigasaki, S., L. Oesterhelweg, A. Gehl, et al. 2004. Application of compact ultrasound imaging device to postmortem diagnosis. Forensic Sci Int 140:33–41. 3. Uchigasaki, S., L. Oesterhelweg, J. P. Sperhake, K. Püschel, and S. Oshida. 2006. Application of ultrasonography to postmortem examination. Diagnosis of pericardial tamponade. Forensic Sci Int 162:167–69.
C.4.3 COMPUTED TOMOGRAPHY (CT) AND MAGNETIC RESONANCE IMAGING (MRI) A detailed overview of the possibilities of postmortem CT is given in the following sections.
C.4.4 MAGNETIC RESONANCE IMAGING (MRI) A detailed overview of the possibilities of postmortem magnetic resonance imaging is given in the following chapters.
157 © 2009 by Taylor & Francis Group, LLC
C5
Documentation of Extracorporeal Findings Silvio Näther, Ursula Buck, and Michael J. Thali
CONTENTS C5.1 Injury-Inflicting Instruments ......................................................................................................................................... 159 C5.2 Incident Scenes .............................................................................................................................................................. 159 C5.2.1 Digital Photogrammetry................................................................................................................................... 159 C5.2.2 Tachymetry ....................................................................................................................................................... 162 C5.2.3 GPS ................................................................................................................................................................... 162 C5.2.4 Terrestrial Laser Scanning ............................................................................................................................... 162 C5.2.5 Application of Surveying Methods................................................................................................................... 164
C5.1 INJURY-INFLICTING INSTRUMENTS In addition to documentation of the autopsy findings of corpses and the injuries of living persons, details of the presumed injury-inflicting instruments must be documented. They can be small objects like teeth, which have bitten someone, and, in strangulation cases for example, a cord or a belt; medium-sized instruments like guns, hammers, or shoes; and large objects|for example, a vehicle in accidents. Also, airplanes and damaged trucks can be documented in 3D. For this approach, optical surface digitizing technology is used. The ATOS II SO system is used for small objects with high demands on accuracy and data quality. In the SO setup, the cameras are put closer to the projector, and objectives of a focal length of 35/50 mm are used. The measurement volumes are 65 mm × 52 mm/35 mm × 28 mm with resulting point spacing of 0.05 mm/0.03 mm and measurement noise amounts up to 0.003 mm/0.001 mm. For small objects it is possible to use TRITOP photogrammetry, but normally it is not necessary. The color information is received from a photograph taken of the object. The objects are fixed in a frame with affixed object points, so it is not necessary to stick points on the object and the optical scanning process can start directly (Figure C5.1.1A,B). For shiny and dark parts, fine white airbrush spray is used to make the surface more suitable for the scanner to capture. The accurate and high-quality result is shown in Figure C5.1.1C. For all other objects the ATOS II and III sensors are used. The ATOS II/III sensor head mounted on a stand is easily positioned around the object or the object is placed on an automatic rotation table, which is normally used for the medium-sized objects. Objects up to 50 kg are placed
in the middle of the rotation table. The rotation table rotates the object being measured automatically in predefined steps. The software carries out these movements automatically and starts the measurements according to the adjusted parameters after each rotation. In this manner the 3D model is built up automatically (Figure C5.1.2). For large objects the optical coordinates measuring system TRITOP is applied at all times. Sometimes 50 to 100 images are taken and automatically computed by the software while photographing. They are scanned with the ATOS III with a measurement volume of 1.5 × 1.5 × 1.5 m. If more details are required the smaller volume of 500 × 500 × 500 mm is used. Vehicles are scanned outside and inside, depending on the needs of the case. In passenger accidents, where a person is run or rolled over, the underbody of the vehicle and the tires have also to be digitized (Figure C5.1.3). A combination of different measuring volumes is possible. Therefore, different sizes of object points are used. Some results of digitized objects are shown in Figure C5.1.4.
C5.2 INCIDENT SCENES Modern technology is also used in the police forensic field.
C5.2.1 DIGITAL PHOTOGRAMMETRY Digital close-range photogrammetry is a useful method for surveying incident scenes. The photogrammetry system can be used for small objects (Chapter B1 in this volume) as well as for rooms and large outdoor areas. The photographs are taken by hand, using a tripod, or from a helicopter or other high position, such as buildings.
159 © 2009 by Taylor & Francis Group, LLC
160
The Virtopsy Approach
A
B
C
FIGURE C5.1.1 (A) Digitalization of the casts of the dentition with the ATOS II SO setup. (B) Fringe pattern projected on the cast. (C) The resolved accurate, high-resolution 3D model of the dentition of the alleged criminal.
A
B
C
FIGURE C5.1.2 (A) Scanning of a pistol with the ATOS III system and an automated rotation table. (B) The 3D model with the slide in the back position and (C) front position.
© 2009 by Taylor & Francis Group, LLC
Documentation of Extracorporeal Findings
161
A
a
C
b
B FIGURE C5.1.3 (A) Digitalization of vehicles with the ATOS III system and a measuring volume of 1.5 × 1.5 × 1.5 meter. (B) The model of the underbody and wheels of a car. (C) 3D model of a deformed car (outside and inside) after a side collision.
FIGURE C5.1.4 Samples of digitized objects: from small parts like a bullet casing, a knife in real color, and a shoe to large complex objects like a damaged motorbike or a crashed helicopter.
© 2009 by Taylor & Francis Group, LLC
162
The Virtopsy Approach
A
B
FIGURE C5.2.5 3D documentation of incident scenes using photogrammetry: (A) Photogrammetric images of a crime scene inside a house, taken from a tripod. (B) 3D model of a crime scene, including the bloodstains and traces, which were based on the photogrammetric analysis, drawn in a CAD program.
The major advantages of the application of photogrammetry in the forensic field are the short preparation time on site and the low budget for the equipment in comparison with other measuring methods. If it is necessary or requested, everything that is photographed can be measured and drawn in a computer-aided design (CAD) program to get a 2D or 3D situation plan (Figure C5.2.5). In combination with digital photogrammetry, additional measurement technologies are used.
C5.2.2 TACHYMETRY With long ranges and a large number of distant traces with high demand on accuracy or when no satellite signals can be received, tachymetry is the best method of choice. Furthermore, in indoor scenes the rooms can be quickly measured with the tachymeter and drawn automatically in 3D CAD software on an Internet-connected laptop. The tachymeter (total station) is a type of theodolite with an electro-optical laser distance-measuring unit (Figure C5.2.6A). It is used in classic geodesy and in engineering surveys to determine the terrestrial or spatial position of points by measuring angles and distances between them. With the modern tachymeter, distances can be measured to a reflector or directly to an object. Using a reflector, the maximum range is 5 km; without a reflector it is 300 m. The maximum measuring distance of the tachymeter used by the Bern police force is 3.5 km with a resolution of 2 mm + 2ppm.
© 2009 by Taylor & Francis Group, LLC
C5.2.3 GPS The global positioning system (GPS) is used for the determination of discrete points in an outdoor incident scene. These points may serve as reference targets to transform the local measurements at the scene into a global coordinate system (georeferencing), such as a Swiss geodetic reference system. The satellite navigation system is used in addition to the tachymetry or the laser scanning for measuring coordinates of discrete points in real time and for measuring distant points (Figure C5.2.6B).
C5.2.4 TERRESTRIAL LASER SCANNING An efficient surveying method that has become more and more important in measurement of complex objects is 3D laser scanning. Such a system allows for the generation of millions of 3D coordinates in a very short time. The 3D laser scanner sends laser beams to the investigation environment while rotating with a horizontal angle of up to 360 degrees and a vertical angle of up to 310 degrees. The scanner measures the distances and angles from the reflected laser beams and computes the 3D coordinates of all reflected objects. A quick digitization of the investigation environment with detailed structures is therefore possible (Figure C5.2.7). Up to 500,000 measuring points per second can be captured in 3D. The maximum range of a terrestrial laser scanner
Documentation of Extracorporeal Findings
163
A
B
FIGURE C5.2.6 Documentation of the scene of an airplane crash using tachymetry and GPS. The black arrows point to the traces and located parts of the crashed airplane; the red arrow points in the flight direction of the airplane: (A) With the total station (tachymeter), angles and distances to points are measured. For example, the position of the traces and parts of the airplane are determined this way. With trigonometry, the angles and distances are used to compute the 3D coordinates of the local positions of the measured points. (B) GPS measurements are used to transform the local measurements at the scene into a global coordinate system. It is also used to determine the coordinates of distantly located parts of the airplane.
A
B
FIGURE C5.2.7 Documentation of a traffic accident scene by using terrestrial laser scanning: (A) The 3D laser scanner acquires threedimensional geometry by sending and receiving laser beams across an object. (B) The laser scanner produces a point cloud of three-dimensional points representing the traffic accident scene, including the wheel traces and damages of facilities (yellow arrows).
© 2009 by Taylor & Francis Group, LLC
164
The Virtopsy Approach
A
B
FIGURE C5.2.8 (A) Point cloud of a house, captured with a 3D laser scanner from outside and inside. (B) View of the rooms of the first floor with all furniture. The second floor and the roof are set invisible.
varies from 50 m to 300 m. These specifications vary with the different products available on the market. These scanners function with two different technologies: the time-offlight method and the phase-shift method. The Bern police force uses a time-of-flight scanner, and the Zürich city police force employs a scanner with the phase-shift measurement principle. The biggest difference between these methods is the range and the speed of the scanners. The result of terrestrial laser scanning is a 3D geo-orientated point cloud with a resolution down to less than 1 × 1 mm. Laser scanning does not depend on ambient illumination. It can be applied in daylight and in darkness, which is a huge advantage for operations in the forensic field.
C5.2.5 APPLICATION OF SURVEYING METHODS These different surveying methods are used for reconstruction of accidents and crime scenes. The currently used method for the documentation of incident scenes is photogrammetry. With this device, 3D true-to-scale drawings of the traces and the replicated scene can be produced by CAD software. These drawings are the basis for further analysis. In road traffic accidents, such analysis includes evaluation of the speed of the vehicles and the accident configuration. In crime scene cases, the 3D drawings are the basis for reconstructions of shooting directions or the analysis of blood spatter.
© 2009 by Taylor & Francis Group, LLC
Additional surveying methods are the GPS system and tachymetry. These systems are used for the determination of 3D coordinates from discrete points. The resulting coordinates are transformed in the global coordinate system and overlaid with maps and terrain models. The advantages are the large area covered and the direct availability of the data for further strategic tasks and analyses. The 3D laser scanner is able to produce a 3D image from the incident scene in the form of point clouds consisting of millions of points (Figure C5.2.8). With a range of 0.1 m up to 300 m, the laser scanner is useful for indoor and outdoor scenes. The scanned data can be combined with the photogrammetric analysis, and the measurements can be taken by tachymetry and GPS. A scene including traces and blood stains is modeled in the 3D CAD program. The produced true-to-scale 3D model of an accident or crime scene is represented in 2D situation plans and 3D views (Figure C5.2.9 and Figure C5.2.10). For the documentation of the exact geometric form, or the deformation, of an involved object at the incident scene, the optical surface digitizing technique is applied. This procedure is useful for a geometric comparison with an inflicting tool or with the injuries of a victim. Figure C5.2.11 shows the photogrammetric documentation of a garage with the TRITOP system. The damaged parts inflicted by a car impact were digitized with the ATOS II optical surface scanner (Figure C5.2.12).
Documentation of Extracorporeal Findings
165
TV
010 009 007
004
040
011
043 041
006
022
012
042 008
FIGURE C5.2.9 Top view of the 3D CAD drawing of the modeled crime scene, including the traces, blood splatter, and the locating position of the victim. The 3D drawing is based on the photogrammetric analysis and on the point cloud resulting from the laser scanning.
FIGURE C5.2.10 3D CAD drawing of the modeled crime scene in side views and a perspective view from a possible position of the offender.
© 2009 by Taylor & Francis Group, LLC
166
The Virtopsy Approach
FIGURE C5.2.11 Three-dimensional documentation of an incident scene using the photogrammetric system GOM TRITOP. The scene is prepared with circular reference targets, coded markers, and coded scale bars. With the help of the coded markers, the calculation of the images and therefore the computation of the 3D coordinates of all markers run fully automatically in no time.
FIGURE C5.2.12 The high-resolution surface models of the damaged door frame, storage rack, and stool, as well as traces on the wall, are the result of 3D surface scanning with a GOM ATOS digitizer. The damages and traces originate from a crash with a car.
© 2009 by Taylor & Francis Group, LLC
Part D Forensic Topics
© 2009 by Taylor & Francis Group, LLC
D1
Radiologic Identification Christian Jackowski and Michael J. Thali
CONTENTS D1.1 Introduction ................................................................................................................................................................... 169 D1.2 CT Imaging of the Dentition ..........................................................................................................................................170 D1.3 CT Imaging of Foreign Material ....................................................................................................................................175 D1.4 Conclusion ..................................................................................................................................................................... 184 References ................................................................................................................................................................................. 184
D1.1 INTRODUCTION The most important diagnosis in everyone’s life is the last one, namely, that the life has ended. Although this statement is based on the correct and efficient medical examinations, the diagnosis of someone’s death is only as correct as the identity of the deceased person who is pronounced dead. Therefore, it is a crucial assignment within the work of forensic pathologists together with crime scene technicians and police officers to ensure the identity of the deceased. In the case of single unknown bodies in daily forensic routine work, or multiple unknown bodies in mass fatality incidents (fortunately less often), to accurately and rapidly identify the victims is important not only for judicial reasons; the correct identity is also required for the next of kin to be able to mourn adequately and thereby accept the death. The International Committee of the Red Cross’s contribution to the 2004 16th meeting of Interpol’s Standing Committee on Disaster Victim Identification states that “identification represents the fulfillment of the right of human beings not to lose their identities after death and, overall, the right of families to know what has happened to their relatives in all circumstances” [1]. With this in mind, the people involved in that challenge can use secure or less secure methods to ascertain the identity of a deceased in forensic practice. Only three techniques are considered secure identification methods within the forensic community, as they do not allow for a reasonable doubt: (1) to compile the individual DNA profile; (2) to use the individual fingerprint; and (3) to compare antemortem and postmortem dental data for identification. All other more frequently used techniques in forensic practice such as confrontation with relatives, recognition of individual body features (e.g., tattoos, piercing, scars), or recognition of personal belongings are less secure, as there are situations when mistakes can occur. Therefore these methods are used when there is already a high grade of assumption concerning the identity, such as when corpses are found in their flats or their cars.
These methods are less elaborate and, of course, less expensive and most of the time do a reasonably good job. However, when using these techniques wrongly positive identification is possible and does sometimes occur. In single unknown corpses there may be one or more assumed identities of missing persons. In these cases the existing antemortem data can be compared to the data of the deceased, and identification can be performed or excluded for the assumed and missing person with reasonable efforts. In contrast, mass fatalities—whether natural disasters (e.g., floods), accidents (e.g., aircraft crashes), or outbreaks of violence (e.g., armed conflicts, acts of terrorism)—are often international in scope so that authorities and experts from several countries are involved in the subsequent workload. Fortunately, an internationally agreed-upon standard exists about procedure for identification of victims of mass disasters: Interpol’s Disaster Victim Identification Guide, which is useful in any type of disaster, regardless of its cause and dimension [2]. The three major stages in victim identification are described: (1) the search for antemortem information of possible victims; (2) the recovery and examination of the bodies to establish postmortem evidence from the deceased; and (3) comparison of antemortem and postmortem data [3]. After every mass disaster, a decision has to be made as to whether or not autopsies should be performed on all victims. While this is undoubtedly advisable, at least in all cases in which a criminal or terrorist connection to the incident is suspected or obvious, it may not always be necessary for the aim of establishing the victims’ identities and causes of death. This would not only be of value for the identification and cause of death aspects but also would assist in preventing or minimizing the effects of similar incidents in the future. Indeed, this appears to be considered standard practice by disaster victim identification (DVI) experts in many countries where mass fatalities have occurred since the mid 1980s [4–8]. However, national law may not deem autopsy of all disaster victims mandatory, and thus even Interpol member states do not always adhere to that recommendation, as illustrated by the 169
© 2009 by Taylor & Francis Group, LLC
170
SAS SK 686 aircraft accident at the Milan Linate airport on October 8, 2001, where the prosecutor in charge ordered judicial autopsies in only 10 of the 118 victims [9]. This example shows that the decision made on this question after a given mass disaster is not simply a matter of the number of victims. It must be noted, though, that all of the mentioned incidents [4–8] relate to catastrophes in which there were no more than 155 victims. So far, no single mass disaster in which Interpol’s recommended DVI procedures have been performed has even remotely approached the dimensions of the December 26, 2004, tsunami disaster in south Asia. For almost as long as they have been in existence, imaging methods have been used not only for diagnostic but also for identification purposes [10]. Routinely used are dental status, evidence of surgical implants, and bony findings such as the sinus frontales or vascular sulci of the cranium. The comparison of antemortem and postmortem roentgenograms is a well-established method in forensic routine. The advent of computed tomography (CT) has given us a further imaging modality to work with. Identification has become possible by comparing antemortem CT scans with postmortem x-ray images [11] and by comparing antemortem and postmortem CT scans [12,13]. More recent advances in CT technology have made the tedious work of producing postmortem x-ray images in which the projections match those of given antmortem images unnecessary by allowing three-dimensional image data to be rotated in virtual space and simulated x-ray images to be calculated [14–16]. This chapter does not claim to serve as an all-embracing summary of all radiological techniques ever applied to compared individual features for identification purposes. Readers further interested in this might consult books such as Forensic Radiology by Gil Brogdon [10]. The chapter rather focuses on the new developments concerning the CT technology to see how these new developments can be used for forensic identification. The chapter should offer insight into current research efforts related to CT-data-based identification and how research efforts might get used in different situations. Therefore, the advances that will support the 3D work are the content of this chapter.
D1.2 CT IMAGING OF THE DENTITION Dental identification plays a major role in identification efforts in mass casualties; for example, the majority of positive identifications of the victims of the 2004 tsunami were at least partly based on the dental characteristics [17]. The teeth, jaws, and orofacial characteristics in general and the specific features of dental work with metallic or composite fillings, crowns, bridges, and removable prostheses as well as distinctive configuration of bony structures of the jaw (mandible and maxilla); the presence and shape of teeth including the roots; the configuration of maxillary sinuses; and long-standing pathology such as prior fractures and orthopedic procedures all can be used [18–22]. Classical methods for forensic dental identification are dental periapical radiographs, bitewing films, and panoramic tomographs.
© 2009 by Taylor & Francis Group, LLC
The Virtopsy Approach
Dental radiographs play a major role in the identification of victims in mass casualties, so the information they contain is invaluable. However, in the field of a mass disaster without having any antemortem x-rays, the main challenge is to document the dental characteristics the best way possible to allow for a comparison when antemortem data have appeared later on. For example, in Thailand after the tsunami postmortem x-ray documentation has been limited to only two images of the bitewings, as dental x-ray documentation of the entire dentition would be inadequately time-consuming under such circumstances. Data acquisition for comparison with any possible antemortem dental radiographs would be useful and could facilitate the work of the forensic odontologists. Multislice computed tomography (MSCT) can serve as an expeditious data acquisition tool. MSCT scans covering the dentition can be performed within 1 to 2 min. As modern CT scanners are available mounted in trailers, it would be possible to get one into the field with minor logistic efforts. The initial reconstructed 2D CT images are of decreased value for identification, but the 3D data distinctively increase the possibilities. One easily applied and easy-to-use possibility is the socalled maximum intensity projection (MIP) as a 3D reconstruction algorithm, which generates a 3D dataset that can be rotated and worked with [16]. The interaction gives the impression to the observer that it is a live x-ray examination that can be influenced in terms of projection, size, or radiation hardness. Having antemortem radiographs to compare, it is possible to rotate, zoom and pan, cut, and whatever else is needed to generate an image that is comparable. Figure D1.1 shows an initially unidentified case out of the Virtopsy Project that has been severely altered due to heat exposure. Comparison of antemortem and postmortem data could be facilitated by using the cranial CT data reconstructed with the MIP mode. Positive identification was possible in this case. The major advantage of this technique is the possibility of adapting the direction of the postmortem dental 3D model to the antemortem radiographs at the time of comparison. Thereby, forensic pathologists or odontologists can choose exactly the same view on the postmortem 3D model as was captured on the antemortem radiograph, thus facilitating the matching of orofacial characteristics in general as well as the specific features of dental work with metallic or composite fillings, bridges, and crowns. Furthermore, they can also assess distinctive configurations of bony structures of the jaw (maxilla and mandible) or the presence and shape of teeth including the roots [18–22]. For mass casualties such as the tsunami in Asia, this would allow for a detailed dental matching with antemortem data (radiographs) at a later point, including the entire dentition and not limited to the two bitewing radiographs. The data acquisition time on the site would be decreased as there is no need to wait for any radiograph to be developed and quality checked until the work in the field can proceed. Otherwise, the advantage of rapid examination time might be realized in the field, as the CT scan would not be limited to the head (see later in this chapter). Furthermore, the quality of the reconstruction might
Radiologic Identification
A
171
B
C
D
FIGURE D1.1 Dental identification using MIP: (A) Initially unidentified burned case that has been CT scanned. (B) MIP of postmortem cranial CT data in that case with window settings at 2,700 HU (w) and 1,800 (c) and an oblique image orientation adapted to the projection of the (C,D) antemortem upper and lower bitewing radiographs that could be used for identification as it is well displaying the dental work that has been performed before.
be increased by adapting the scanning parameter with a collimation of 0.5 mm or lower, resulting in isotropic voxels as well as in a thinner slice thickness and an increased increment for at least the cranial CT data, which would cost another 1 to 2 min. A further advantage is the noninvasiveness of the documentation, which makes jaw resection or forcible opening of the mouth unnecessary. Thereby, especially the fragile teeth in burned corpses can be documented without being touched, thus avoiding the danger of destroying the brittle dentition. The acquired data exist in a digital form and can be sent online to any forensic odontologist or treating dentist within minutes. Finally, the possibility of rearranging bony and dental fragments in comminuted fractures of the jaw within a 3D model of the dental CT of trauma cases data should be mentioned, as it allows for a better comparison with antemortem radiographs of the intact jaws. The major disadvantage currently of just applying the MIP on normal cranial CT data may be streak artifacts caused by an abundance of metallic dental work that might decrease the quality of 3D reconstructions in certain cases. The detailed documentation of the shape of the restorations can thereby become less precise compared with classical radiographs. However, at least one of the major causations for streak artifacts—namely, the amalgam—is increasingly being replaced by more radiolucent nonmetallic composite materials [23]. Furthermore, removable prostheses containing metallic material can be removed prior to scanning if the quality is intolerably decreased. Another way to use the CT data to support identification based on dentition is to generate multiple panoramic images. Dentascan software is used in the clinical environment to evaluate dental implants and to assess tumors, cysts, inflammatory diseases, fractures, and surgical procedures [21]. As more and
© 2009 by Taylor & Francis Group, LLC
more antemortem panoramic images are created using x-ray or CT technology, there are increasing panoramic images available for identification purposes [24]. Thereby the Dentascan software applied on postmortem CT data of the dentition gains more importance for dental identification, as this is a feasible way to get comparable images (Figure D1.2, Figure D1.3, and Figure D1.4). There is a complete documentation of the entire dentition and the jaws. The main limitation of the Dentascan images is that they sometimes provide poor visualization of the shape of the fillings. Otherwise, it gives a quick overview in a very user-friendly way of the dentition (missing or nonerupted teeth) and the location of fillings, which can easily rule out any assumed identity. Because of its weakness concerning visualization of the shape of restorations, we do not suggest using Dentascan images alone for a positive identification if there is no further supporting finding or hint. The most promising approach for future applications goes along with a distinctive decrease of streak artifacts and offers the opportunity to even color encode the used restorative materials within 3D volume-rendered models. The secret to this is to use a special algorithm called extended CT scale to calculate the initial CT images. Normally there are only physiological body tissue radiopacities included within the CT images, and everything that shows distinctive higher radiopacities causes artifacts. The extended CT scale allows widening the spectrum of radiopacities to be included up to 10 times. The normal range is somewhere between –1,000 and 3,000 to 4,000 HU depending on the hardware supplier. (Some hardware suppliers have their upper limit at 3000 HU [General Electric]; others such as Siemens have it at 4000 HUs. The lower limit is always –1000 HU.) Using the extended scale increases this range up to –10,000 to 30,000 to 40,000 HU. Thereby radiopacities that some lighter metals such as aluminum or steel have are reached. Heavier metals such as
172
The Virtopsy Approach
A
C
B
D
FIGURE D1.2 Workflow using the Dentascan processing software: (A) On a lateral view of a 3D transparent MIP model of the skull (as also seen in Figure D1.1B), a level for an oblique section has to be manually defined that shows the entire dentition in one cross-sectional image (demonstrated in (B)) by using the red line within the image. (B) On that image the position of the teeth has to be manually appointed. (C) Based on that information the software generates curved lines defining new curved reformatted sections along the dentition. (D) Along those lines parallel sections in a user-defined number and thickness can be calculated as a panoramic dental image for comparison with antemortem dental radiographs.
A
B
FIGURE D1.3 Dental profiling with Dentascan images: (A) Thin postmortem panoramic Dentascan image. (B) Bitewing radiographs to compare. Note the analogy in teeth anatomy (e.g., the missing of all wisdom teeth) and the number and location of dental fillings. Otherwise, the detailed shape of the fillings cannot be assessed on the generated Dentascan image.
© 2009 by Taylor & Francis Group, LLC
Radiologic Identification
173
FIGURE D1.4 Dental profiling with Dentascan images in two badly burned corpses. Although the dentition was altered during the exposure to heat, the postmortem Dentascan images (middle) could display dental characteristics that allowed for a positive dental profiling in both of the cases. In the right one, additional postmortem radiographs were obtained to ensure the victim’s identity.
lead or the filling material amalgam are still beyond these limits, but a lot of materials that are used within the odontological practice are represented within extended scale images without artifacts. Otherwise, knowledge about the CT behavior of these materials is still insufficient for routine use. The first study addressing this new forensic field was a feasibility study. In close collaboration with the Department of Operative, Preventive and Paediatric Dentistry, we scanned extracted teeth presenting different fillings using our clinical 6 detector row MSCT scanner (Emotion 6, Siemens Medical, Erlangen, Germany) [25]. The roots of the teeth were embedded in Historesin as a mounting medium, and the scans were performed with the teeth immersed in artificial saliva [26] to simulate oral conditions. Scanning parameters were as follows: raw data acquisition (130 kv, 280 mAs, 0.5 mm collimation, rot. time 1.5 s, pitch 1.6) and image reconstruction (0.625 mm thickness, 0.1 mm increment, B30 reconstruction kernel, matrix 512, field of view [FOV] 50 mm, extended CT scale). The radiopacities of the filling materials were analyzed in coronal 2D images on a workstation (Leonardo, Siemens Medical, Erlangen, Germany) (Figure D1.5). The initially obtained values varied in wide ranges [25], and it was retrospectively not possible to determine exactly what material was used in terms of company and composition. Therefore, it seemed not to be reasonable to create a color-encoded Hounsfield unit (HU) ramp for each material as initially intended, which is why the values were used only to discriminate between amalgam and further metallic materials such as gold and temporary and definite composite filling materials. But this knowledge could already be implemented as a volumerendering preset (Inspace, Syngo, Siemens Medical, Erlangen,
© 2009 by Taylor & Francis Group, LLC
Germany) that color encodes the different material groups easily visualized within reconstructed 3D volume-rendered 3D models of the dentition [25]. An HU ramp for the volumerendering software was defined for temporary and composite fillings (4,500–17,000 colored red) and for amalgam as well as more radiopaque materials such as gold (19,500 – 30,710 HU colored green). Additionally, a ramp was enclosed to encode enamel and less radiopaque dental structures (1,000 – 4,500 HU) to visualize the teeth within the volume. The resulting software application was applied to postmortem dental CT data of forensic cases and proved itself well (Figure D1.6). The postprocessing algorithm was able to roughly discriminate the composition of dental fillings in a 3D volumerendered model of the dentition. Application of the created preset allows for a very quick overview of the dental restorations of a corpse concerning the two described groups. Thereby not only the location, size, and shape of the fillings but also the used material could be compared with dental records of the treating dentist. This information is obtained within minutes after a postmortem CT scan. A minor disadvantage was that the fillings with increased radiopacity (e.g., amalgam) were always displayed with a small margin that encodes the material with lower radiopacities within the reconstructed 3D model. This is in part caused by partial volume effects at the border of the filling itself. The distinctively reduced but still partly remaining streak artifacts within the initial 2D images contribute to this phenomenon. There is a continuous decrease in HU at the border between filling and tooth or oral cavity (see graph in Figure D1.5) also passing the specific HUs of lower radiopacity filling materials. This is thereby a normal finding
174
The Virtopsy Approach
A
B
7400
1 Distance: 1.13 cm 1 Min/Max: 30/7400 5558 1
3715
1873
30 0.0 0.6 1.1 cm Position on Line 1
FIGURE D1.5 Setting of a study that showed color encoding of different filling materials within a reconstructed 3D CT model: (A) Extracted human tooth with filling that underwent a CT examination. The blue dotted line indicates the cross-section that is demonstrated in B. (B) Cross-section through the tooth clearly depicts the filling. Measurement of the CT radiopacity in HU along the line indicated with 1 gives the graph seen on the right. Note two peaks at 3,700 HU caused by the enamel and the peak of the temporary filling material at about 7,400 HU.
A
B
FIGURE D1.6 Color encoding of filling materials using CT data: (A) Volume rendering preset to color encode amalgam and metallic filling materials (e.g., gold as light green), separating temporary and composite materials colored as red based on the information obtained on the extracted teeth (see Figure D1.5). Bony structures of the jaws and parts of the teeth are represented by white voxels. (B) Forensic example case with different composite fillings (red) and one amalgam filling (light green).
© 2009 by Taylor & Francis Group, LLC
Radiologic Identification
and does not display a composition of different filling materials. If a detailed shape of a filling is aimed for, it can be achieved by turning off the HU ramp of the lower radiopacity material (Figure D1.7). Only the homogeneous central color of the filling encodes its composition with certainty. However, the lowest HUs chosen cover the enamel and only parts of the bony structures of the jaw, which also reduces artifacts within the model. The material that might cause the major challenge for CT documentation is ceramic. It showed similar or even lower radiopacity values compared with the teeth. That distinctively complicates its discrimination within the CT data, as it results in a similar white visualization of the ceramic material when visualized using the mentioned preset. However, as our investigated ceramic crown also consisted of metal parts, its detection should be certain as, for example, further metallic can be nicely represented within the dental images (Figure D1.8). A minor disadvantage may be the lack in discrimination between metals of more than 30,710 HU (e.g., amalgam,
175
gold). This might become possible when CT systems are implemented that have a further increased extended CT scale than is presently available. Otherwise, it is not unlikely that CT systems will become implemented that are adapted for postmortem use. That would allow providing CT scanners with x-ray tubes that are distinctively more powerful, as there is no concern about the applied radiation dose postmortem. Comparable scanners already exist to perform material analysis for industrial purposes. Nevertheless, the era of streak artifact caused impossible assessment of the restoration status of the dentition based on postmortem CT data seemed to be over.
D1.3 CT IMAGING OF FOREIGN MATERIAL Foreign material is of great value for identification. Since today many osteosynthesis and endoprosthesis techniques are widespread, in their application a lot of foreign material can be found in unknown corpses that is useful for identification purposes. Normally these techniques also have antemortem
FIGURE D1.7 Shape of fillings. By knowing the radiopacity of the individual filling material, the VR preset can be improved until the detailed shape of the fillings become represented as demonstrated in two examples (amalgam upper images; temporary filling material lower images).
© 2009 by Taylor & Francis Group, LLC
176
The Virtopsy Approach
A
B
FIGURE D1.8 Additional dental findings that are useful for identification purposes documented by postmortem CT. In these cases metallic wires became obvious within the CT data that were attached to the frontal teeth to push them in line.
x-ray documentation. But from plain x-ray documentation to 3D computed tomography documentation of foreign bodies, an enormous advantage is becoming obvious. As already mentioned, for the dental CT data, 3D computed tomography allows for an adaptation of the postmortem images to any antemortem x-ray projection that becomes available for identification. This is less important in single cases that come to a forensic institution with assumed identity that needs to be proven or ruled out. CD computed tomography gains more importance in mass fatalities when there are no antemortem radiographs available at the time of postmortem
data acquisition, such as the tsunami catastrophe in Asia in December 2004 [17]. Postmortem CT offers the chance of working with the data as a complete 3D dataset of the corpse, which provides the opportunity to zoom in and out, turn around, or cut within the model. According to our experience so far with the routine application of CT scanning in forensic case work, a great deal of valuable information useful for identification can be provided by the CT data. During the CT scans that were performed for different forensic questions, we recognized a lot of incidental findings that helped to ensure the identity of the corpses (Figure D1.9, Figure D1.10,
A
B
C
D
E
F
FIGURE D1.9 Incidental postmortem CT findings supporting the identification: foreign bodies—part one: (A) Helical wire within the left humerus after fracture. (B) Screws within the right humerus after fracture. (C) Screw within the dens of the axis after fracture. (D) Screw within the tibia. (E) Two screws within the left femur. (F) Screws within the right femur and tibia after surgical treatment of a rupture of a cruciate ligament.
© 2009 by Taylor & Francis Group, LLC
Radiologic Identification
177
A
B
FIGURE D1.10 Incidental postmortem CT findings supporting the identification: foreign bodies—part two: (A) Endoprosthesis of the right humerus. (B) Total endoprosthesis of the right hip.
Figure D1.11, Figure D1.12, Figure D1.13, Figure D1.14, and Figure D1.15). Experienced users can also draw conclusions to some extent from the radiopacity data on the component on the material that was used. In a very self-critical way we had to reconceive the routine procedures used thus far, as most of the incidental findings would not become obvious in a routine autopsy. Otherwise, performing plain radiographs on the entire corpse, which
A
would also demonstrate many of these findings, would be much more time-consuming and elaborate. In early spring 2005—being back in Switzerland from Thailand and the initial DVI forensic postmortem case work and thereby recently having experienced the problems and challenges of DVI management in the aftermath of an incident with dimensions as caused by the tsunami—we felt, based on our CT experiences, that the technology could distinctively
B
C
D
E
FIGURE D1.11 Incidental postmortem CT findings supporting the identification: foreign bodies—part three: (A) Surgical plate on the left ulna after fracture. (B) Surgical plate on the lumbar spine. (C) Surgical plate on the left tibia after fracture. (D) Surgical plate on the cervical spine after fracture. (E) Surgical plate on the right femur.
© 2009 by Taylor & Francis Group, LLC
178
The Virtopsy Approach
A
B
C
FIGURE D1.12 Incidental postmortem CT findings supporting the identification: foreign bodies—part four: (A–C) Three cases with minimal invasive treatment of osteoporosis, the so-called vertebroplasty, that causes a very individual shape of bone cement within the human spine.
support the postmortem data acquisition of the DVI teams in the field. To scientifically address this idea, a study was initiated to theoretically test postmortem CT data to complete the documents routinely done on the field: Interpol’s postmortem form [27]. We analyzed Interpol’s pink postmortem (PM) form in order to identify parts that could possibly be filled in with the information gathered from a victim’s postmortem CT data
A
rather than from direct observation of the victim’s dead body. The antemortem (AM) and PM forms are divided into seven sections, A–G, as follows: (A) Personal data of the missing person (not present in the PM form) (B) Recovery of the body from the site (not present in the AM form)
B
FIGURE D1.13 Incidental postmortem CT findings supporting the identification: foreign bodies—part five: (A) Elaborate surgical treatment of a hip fracture in an initially unidentified victim of an airplane crash (note the various fracture systems). (B) Magnification even reveals the shape of the foreign material in a very detailed way.
© 2009 by Taylor & Francis Group, LLC
Radiologic Identification
179
A
B
C
FIGURE D1.14 Incidental postmortem CT findings supporting the identification: foreign bodies—part six: (A) Antemortem radiograph showing osteosynthesis of right tibia by medullary nailing. (B, C) Postmortem reformatted CT images show the canal that remained after removal of the nail and the smaller horizontal canals caused by its fixating screws.
A
B
C
(C) Description of effects: (1) clothing and shoes; (2) personal effects; (3) jewelry (D) Physical description of the body, distinguishing marks (e.g., tattoos) (E) Medical information about the missing person (AM); data obtained by internal examination of the body (PM) (F) Dental information (AM); dental data (PM) (G) Any further information that may assist in identification On the PM form, these sections are usually filled in by the following persons: (B) (C) (D) (E) (F) (G)
FIGURE D1.15 Anatomical findings useful for identification: (a) Postmortem CT scan of a badly decomposed body revealed some local intra-abdominal calcification (arrow). (b,c) Ordered antemortem abdominal radiographs show the same finding diagnosed as a calcified lymph node within the medical records. Nevertheless the finding could help to ensure the victim’s identity.
© 2009 by Taylor & Francis Group, LLC
Police, fire brigade, civil protection, military, other Police, fire brigade, civil protection, military, other Forensic pathologist Forensic pathologist Forensic odontologist Various persons
We concentrated on sections D and E of the PM form, as these are the ones to be filled in by the forensic pathologist and, together with section F, the ones most likely to accept data obtained from CT scans. All items pertaining to color (e.g., skin, eyes) were excluded a priori, as CT images do not give real color information. Equally, all items concerning hair
180
The Virtopsy Approach
(even if not pertaining to its color) were excluded because the diameter of a single hair is less than 0.625 mm, which is the resolution limit of the CT scanner used. We postulated all other items except circumcision status (item D3-54) to be suitable for documentation by CT. A list of those items from sections D and E postulated to be suitable for documentation by CT data was compiled. Postmortem CT scans were then performed on 32 forensic routine cases with a generated slice thickness of 1.25 mm. We then investigated whether the information required for the proposed items of the Interpol PM form can be gained from the CT data. Table D1.1 gives a detailed overview of the contents and displays the results of that study. Beyond what is listed in Table D1.1, pages D4 and E3 can also be completed with CT findings. They consist of a body sketch and a skeleton sketch, which are completed with details already documented in items D1-31 (“State of the body”), D3-53 (“Specific details”), and El-64 (“Skeleton/Soft tissue” under the heading “Internal examination—Full autopsy”). In item D1-31, which describes the state of the body, the forensic pathologist first has to record whether the body is visually identifiable or not. The actual description that follows can then be based on CT images. Twelve body regions—head, neck/throat, right arm, left arm, right hand, left hand, body front, body back, right leg, left leg, right foot, and left foot—are separately described as damaged, burned, decomposed, skeletonized, missing, or loose; CT has been proven useful to document this. It has turned out to be particularly good at localizing both intensity and direction of heat in burned bodies [28]. To establish the height of the body (item D1-32), one of several methods may be used. With a continuous set of CT data, virtual sections at arbitrary angles through the volume can be calculated; then the distance between two points|such as both ends of a bone|within such a plane can be measured. If the body can be laid outstretched on the examination table, its full length can even be determined from the CT data in one single measurement. If this is not possible due to contractures (e.g., in burned victims), the body height can still be determined by adding the measured lengths of a number of skeleton segments. Finally, if the body is not complete, one of
TABLE D1.1 Relevant Items in Sections D and E of Interpol’s DVI PM Form a b c d e f g h i j
It must be recorded whether the body is visually identifiable Age estimation based on CT data is possible but not preferred Weight can only be estimated from the body volume measured Spectacle marks, if shallow, may not be visible in CT Lip makeup is not visible in CT Denture ID number is not visible in CT Nail paint is not visible in CT Nicotine stains are not visible in CT Tattoos are not visible in CT Samples may be taken as CT-guided postmortem biopsies
© 2009 by Taylor & Francis Group, LLC
several anthropological methods can be applied [29]. These are expressed as formulas that have the lengths of several of the long bones of the extremities as parameters [30,31]. Figure D1.16 shows measurements of a badly decomposed victim’s humerus, radius, femur, and tibia in reformatted CT images. The weight of a body (item D1-33) is best established by weighing the body, but since body volume, which can be measured from the data of a full body CT scan, can also be correlated to the body weight, there is an excellent correlation of r = 0.9966 [32]. Thereby it is at least possible to estimate the body weight using CT data if doubts arise later concerning the documented weight or if weighing has been missed altogether. In items Dl-34–D3-54, the forensic pathologist is asked for physical descriptions of various parts of the body based on external inspection. Among these, about 60% can at least partially be described using 3D CT data reconstructions of soft tissues (Table D1.1). A relatively small number of reconstructed images can be used for many different items. In item D1-34 (build), the forensic pathologist is asked for a description of a victim’s bodily constitution (light, medium, or heavy; item 34.01), head form in the frontal view (oval, pointheaded, pyramidal, circular, rectangular, or quadrangular; item 34.02), and head form in the profile (shallow, medium, or deep; item 34.03). While even unaltered CT slices allow item 34.01 to be established, the bodily constitution and three-dimensional reconstructions of the soft tissues of the head must be calculated so as to establish items 34.02 and 34.03 (the head form) as if by looking at the actual body. Figure D1.17 shows the head of one of the scanned forensic cases, both in the frontal view and in profile, in 3D volumerendered CT data, and in photographic images for comparison. When deciding on items 34.02 and 34.03, it is advisable to consult the silhouette sketches that are provided (at the end of the comparison report form). Item D1-35 (race) asks a victim’s race to be determined as caucasoid, mongoloid, or negroid and the complexion as light, medium, or dark. Since various skeletal features are used by anthropologists for race determination, we expect that 3D reconstructions of bones may allow this task to be completed without the need to skeletonize the bones. In item D2-37 a body’s forehead has to be described in terms of height/width (low, medium, or high; narrow, medium, or wide; item 37.01) and inclination (protruding, vertical, slightly receding, or clearly receding; item 37.02). This can be done with the help of the reconstructed images that have already been calculated for item D1-34. In item D2-40, the nose is to be described in terms of size/shape (small, medium, or large; pointed, roman [i.e., with a prominent bridge], or alcoholic; item 40.01), presence or absence of spectacle marks (and other peculiarities; item 40.02), and curve/angle (concave, straight, or convex; turned down, horizontal, or turned up; item 40.03). With the possible exception of spectacle marks—namely, if they are too shallow—these descriptions too are feasible using the same images as before. As an aid in deciding on item 40.03 (curve/angle of the nose),
Radiologic Identification
181
A
B
340.5 mm (3D)
C
249.8 mm (3D)
D
388.0 mm (2D) 463.7 mm (3D)
FIGURE D1.16 Estimation of body height. Distance measurements within reformatted images that have been chosen containing both end points of a selected bone; these measures can be used for an estimation of the body height in a badly decomposed or burned corpse as well as when only body parts are remaining: (A) humerus, (B) radius, (C) femur, (D) tibia.
silhouette sketches similar to those of the head form are provided. Item D2-42 asks for a description of the ears in terms of size/angle (small, medium, or large; close-set, medium, or protruding; item 42.01) and of earlobe attachment and piercing (item 42.02). Silhouette sketches of earlobe attachment are provided. Items D2-43 and D2-44 describe the mouth in terms of size (small, medium, or large) and the lips in terms of shape (thin, medium, or thick). The presence of lip makeup, which cannot be seen in any kind of CT images, is recorded in item 43 also. Items D3-47 and D3-48 describe the chin in terms of size/ inclination (small, medium, or large; receding, medium, or protruding; item 47.01) and shape (pointed, round, angular,
© 2009 by Taylor & Francis Group, LLC
cleft chin, or groove; item 47.02) and the neck in terms of length/shape (short, medium, or long; thin, medium, or thick; item 48.01) and peculiarities (goiter, prominent Adam’s apple, collar/shirt number, circumference in cm). No new images need to be computed for item 47, as the needed information is found in the ones used for items 34.02 and 34.03 (the head form). In items D3-49 and D3-50, the hands and feet are described in terms of shape/size, nails, and peculiarities. While it is impossible to perceive nail paint or nicotine stains in any kind of CT images, width and length of hands and feet and length of nails can be described with the help of threedimensional reconstructions. It is evident that the descriptions given in the aforementioned items are subjective in character; different examiners
182
A
The Virtopsy Approach
B
34
02 Head form, front (Shape of head from front)
1 Oval
2 Pointheaded
3 Pyramidal
4 Circular
5 Rectangular
6 Quadrangular
03 Head form, profile (Shape of head from side)
C
D 1 Shallow
2 Medium
3 Deep
a = Data not available/indefinable
b = Photo
c = Further information on page 6 a b c
Physical Description (at mortuary) 34
Build
Light
01 Bodily constitution
2
1 Oval
02 Head form, front ?? 03 Head form, profile
Medium
Pointheaded
1
2
Shallow
Medium
1
2
Heavy 3 Pyramidal Circular 3
4
Rectangular 5
Quadrangular 6
Deep 3
FIGURE D1.17 Three-dimensional data storage of the victim’s shape: (A,B) Postmortem routinely obtained photographs for documentation are compared to a (C,D) volume-rendered 3D visualization of the postmortem CT data. Note that the shape of every detail contributing to the individual aspect is documented in 3D to scale. That would allow retrospective measurements to be taken such as the head circumference to estimate the hat size or the neck circumference to estimate the collar size of shirts; glasses that are suspected belongings may even be digitally tried on. Note that the only features not also visible in the 3D reconstructions are single hair and color of hair and skin. Otherwise, the data can be used for compilation or verification of Interpol PM form items (examples from the document are given on the right) such as Dl-34.02 (head form, front), D2-37.01 (forehead: height/width), D2-42.01 (ears: size/angle), D2-43 (mouth), D2-44 (lips), D2-47.02 (chin: shape), Dl-34.03 (head form, profile), D2-37.02 (forehead: inclination), D2-40 (nose), D2-42.01 (ears: size/angle), or D2-74.01 (chin: size/ inclination).
might describe the same person differently. Figure D1.17 illustrates this difficulty. One may try to decide whether the head in this given case is oval or pointheaded and whether it is medium or deep in profile. Such decisions are even more difficult in those items where no sketch is provided, as for nose or ear size. There is always the risk that an examiner may consider noses and ears to be of medium size when they are similar to his or her own nose and ears; this can make matching AM and PM forms difficult, as they are virtually never compiled by the same person. It must be noted, though, that these difficulties are the same whether one inspects an actual dead body or three-dimensional CT reconstructions. The advantage of the CT approach lies in documentation. Whereas photographic documentation limits the views of a deceased to a number of angles, from a CT dataset reconstructed views from different angles can be calculated in an infinite number, at any time. If, for example, no picture has been taken showing the ears’ angle, what the examiner wrote down in item D2-42.01 can still be controlled later if CT data of the head are available. It may also be seen as an advantage that reconstructed images do not have the same emotional impact on the person who decides on whether the identity is correct or not. Using item D2-45, the forensic pathologist has to roughly describe a victim’s teeth in terms of condition (natural, untreated, treated, crowns, bridges, or implants; item 45.01), gaps and missing teeth (item 45.02), and dentures (part, upper,
© 2009 by Taylor & Francis Group, LLC
part, lower, full upper, full lower, ID number; item 45.03). A much more thorough listing of dental findings is required in section F of the form, which is to be filled in by a forensically trained odontologist. Both tasks can be accomplished by dental CT data as already discussed herein. Item 53 allows a record to be taken of specific details of the same 12 regions of the body as item 31 (the state of the body), such as scars/piercings, skin marks, tattoo marks, malformations, and amputations, some of which are clearly distinguishable in CT images, and item 55 allows other peculiarities to be documented. In order to complete the PM form, before or after the CT examination the forensic pathologist should systematically do the following: r Inspect the entire body, and specify race and complexion (item 35), describe body hair and pubic hair (in terms of extent and color; items 51 and 52), and record tattoos, scars, burns, and so forth (items 31 and 53). r Check the body from top to toe, and describe the hair of the head (in terms of, e.g., type, length, color, shade of color, thickness, style, baldness; item 36), the eyebrows (item 38), eyes (item 39), and facial hair (item 41); check the nose for spectacle marks (item 40.02) and the lips for makeup (item 44.01); record the ID number of dentures if
Radiologic Identification
present (item 45.03); check for nicotine stains on teeth, lips, and fingers (items 46.01 and 49.03) and for nail makeup on fingers and toes (items 49.03 and 50.02). r If male, check circumcision status (item 54). r Estimate the individual’s age (item 31 A). r If possible, measure the body height (item 32) and weigh the body (item 33). Carrying out these steps ensures that items of the PM form that are not suitable for documentation by CT do not remain blank. Items El-60–El-65 are meant for recording findings of an internal examination (i.e., a full autopsy). The goals, however, are not the same ones as a pathologist performing a regular autopsy. In the DVI context the concentration is mainly on finding clues for the victim’s identification and secondarily on the cause of death. If full-body CT scans have been performed on all dead bodies instead of performing autopsies, the CT data can be used to perform a virtual autopsy. In this way, many features that can be helpful for the purpose of identification of a body may be found that otherwise might never be detected, as the CT examination may even reveal clues for identification that would remain hidden to the eye during autopsy (Figure D1.9, Figure D1.10, Figure D1.11, Figure D1.12, Figure D1.13, Figure D1.14, and Figure D1.15). The decision about a victim’s sex (item E2-71) is usually based on inspection of external (primary and secondary) sexual characteristics, but the whole-body CT scan also shows internal primary sexual characteristics. Three-dimensional reconstructions of a scanned victim’s skeleton even make it possible to apply anthropological methods of sex determination to skeletal features if necessary, without the need for time-consuming preparation of bones (Figure D1.18). We concluded from these results that MSCT can be integrated as a valuable screening tool into the DVI process that has been proposed by Interpol. It does not make any other step in the process unnecessary, yet extending that process by one more step can help reduce the time required at the site to acquire the postmortem data needed to establish the victims’
A
183
identities. Performing CT scans of all victims could prove especially helpful (a) when there is a large number of victims, and certain circumstances such as warm and humid climate and lack of cooled storage room threaten to prevent timely autopsy of all of them, and (b) when performing autopsies, even for medicolegal purposes, is not socially accepted at the place of the disaster by religious or other reasons. A full-body CT scan only takes about 10–15 min per body (e.g., the positioning), theoretically allowing 4–5 bodies to be scanned per hour; thus, calculating with unforeseeable work pauses, scanning 80 bodies per 24 h appears to be realistic. Since the scan acquires data that contain about 60% of the information needed for the extensive physical description in section D of the Interpol PM form [27], time needed per corpse in the field is drastically reduced. Reformatting and interpreting one victim’s CT data can be done or even postponed while bodies continue to be scanned or until all bodies have been scanned, depending on the available computer hardware, number of personnel available, and dimensions of the disaster. Theoretically, at a rate of 80 bodies per 24 h, three forensically trained radiologists or radiologically trained forensic pathologists working simultaneously could gather all the information that is needed but if interpretation is postponed, as we suggest, only one radiologist would be needed on site for quality management. The MSCT model used in this study was an air-cooled system, necessitating pauses during full-body scans of adequate collimation (e.g., 1 mm or less) for the tube to cool down. The water-cooled straton technology, in contrast, would allow exploiting the full potential of MSCT in the field, even with a victim number comparable to that of the 2004 tsunami. Other than efficient cooling, to fulfill the needs of a DVI team a CT scanner should have an extended FOV so as to make images of, for example, burned victims in fencer’s posture possible, and an extended CT scale, which is needed to characterize different materials of high radiopacity such as dental filling materials [25]. Even under conditions better than those described, performing CT scans in the course of the DVI process has advantages not reached by any other method. It allows
B
FIGURE D1.18 Estimation of sex: (A,B) AP- and cranio-caudal view on a volume-rendered bony model of the pelvis, clearly depicting male characteristics such as the acute subpubic angle as well as the heart-shaped pelvic aperture. No maceration of the bones to investigate is necessary.
© 2009 by Taylor & Francis Group, LLC
184
The Virtopsy Approach
FIGURE D1.19 A mobile CT in the field. As modern CT scanners have already been mounted on trailers to become a mobile diagnostic tool, it would be worth it to implement their use within the work flow of disaster victim identification teams. Data acquisition is rapid, and the unidentified corpses could be stored as a real data-based 3D data set in their actual condition to be consulted at any time for any questions CT data do not putrefy.
observer-independent, objective documentation even of those items in the PM form that are not strictly quantifiable, such as nose size (D2-40.01) or lip shape (D2-44.01). Any foreign body, be it small like a coronary stent or big like an arthroplasty, will be detected and can contribute identification. If antemortem information does not become available until a long time after the disaster, identification may still be possible, providing the CT data have been stored. Otherwise, if the initial examinations have been done without the use of CT it may be used for reexamination to gain more information. The CT data can be distributed to forensic pathologists in suspected victims’ home countries electronically, so the work can be done in a decentralized manner in geographically distant places (what could be called teleforensics), and the number of specialists needed at the actual disaster site can be reduced, contributing to a decrease in the overall logistic efforts.
D1.4 CONCLUSION In the view of the experiences gained so far within the Virtopsy Project, the apparent increase of both frequency and scope of mass disasters, and the increasing availability, affordability, and ease-of-use of MSCT equipment, the CT technology has to be introduced into the DVI process as a screening tool (Figure D1.19). More than half of the information needed to
© 2009 by Taylor & Francis Group, LLC
fill in the postmortem form can be documented in a nondestructive, three-dimensional, and objective manner. It does not make any other step in the process unnecessary, yet extending that process by one more step can help to reduce the time needed to establish victims’ identities. We are convinced that MSCT will gain more importance for the identification of corpses when its technology becomes implemented at forensic institutes or as mobile CTs supporting the DVI teams in the field and when it is appreciated as a rapid tool for data acquisition that offers new possibilities to compare antemortem and postmortem individual characteristics.
REFERENCES 1. International Committee of the Red Cross. 2007. The handling of human remains and information on the dead in situations relating to armed conflicts or internal violence and involving missing persons. http://www.icrc.org/Web/ eng/siteeng0.nsf/htmlall/5ZMJH5/$File/Interpol_ 2004DVI_EN.pdf. 2. International Criminal Police Organization. 1997. Disaster victim identification guide. http://www.interpol.int/Public/ DisasterVictim/guide/default.asp. 3. Pan American Health Organization. Management of dead bodies in disaster situations. http://www.paho.org/english/ dd/ped/DeadBodiesBook.pdf. 2004.
Radiologic Identification
4. Meyer, H. J. 2003. The Kaprun cable car fire disaster—aspects of forensic organisation following a mass fatality with 155 victims. Forensic Sci Int 138:1–7. 5. Soomer, H., H. Ranta, and A. Penttila. 2001. Identification of victims from the M/S Estonia. Int J Legal Med 114: 259–62. 6. Hutt, J. M., B. Ludes, B. Kaess, A. Tracqui, and P. Mangin. 1995. Odontological identification of the victims of flight AI.IT 5148 air disaster Lyon-Strasbourg 20.01.1992. Int J Legal Med 107:275–79. 7. Timperman, J. 1991. How some medicolegal aspects of the Zeebrugge Ferry disaster apply to the investigation of mass disasters. Am J Forensic Med Pathol 12:286–90. 8. McCarty, V. O., A. P. Sohn, R. S. Ritzlin, and J. H. Gauthier. 1987. Scene investigation, identification, and victim examination following the accident of Galaxy 203: disaster preplanning does work. J Forensic Sci 32:983–87. 9. Lunetta, P., H. Ranta, C. Cattaneo, et al. 2003. International collaboration in mass disasters involving foreign nationals within the EU: medico-legal investigation of Finnish victims of the Milan Linate airport SAS SK 686 aircraft accident on 8 October 2001. Int J Legal Med 117:204–10. 10. Brogdon, B. G. 1998. Forensic Radiology. Boca Raton, FL: CRC Press. 11. Haglund, W.D. and C. L. Fligner. 1993. Confirmation of human identification using computerized tomography (CT). J Forensic Sci 38:708–12. 12. Smith, D. R., K. G. Limbird, and J. M. Hoffman. 2002. Identification of human skeletal remains by comparison of bony details of the cranium using computerized tomographic (CT) scans. J Forensic Sci 47:937–39. 13. Reichs, K. J. 1993. Quantified comparison of frontal sinus patterns by means of computed tomography. Forensic Sci Int 61:141–68. 14. Riepert, T., D. Ulmcke, F. Schweden, and B. Nafe. 2001. Identification of unknown dead bodies by x-ray image comparison of the skull using the x-ray simulation program FoXSIS. Forensic Sci Int 117:89–98. 15. Riepert, T., D. Ulmcke, U. Jendrysiak, and C. Rittner. 1995. Computer-assisted simulation of conventional roentgenograms from three-dimensional computed tomography (CT) data–an aid in the identification of unknown corpses (FoXSIS). Forensic Sci Int 71:199–204. 16. Jackowski, C., E. Aghayev, M. Sonnenschein, R. Dirnhofer, and M. J. Thali. 2005. Maximum intensity projection of cranial computed tomography data for dental identification. Int J Legal Med 120:1165–67. 17. James, H. 2005. Thai tsunami victim identification overview to date. J Forensic Odontostomatol 23:1–18.
© 2009 by Taylor & Francis Group, LLC
185
18. Fixott, R. H. 2001. How to become involved in forensic odontology. Dent Clin North Am 45:417–25. 19. Fixott, R. H., D. Arendt, B. Chrz, J Filippi, J. McGivney, and A. Warnick. 2001. Role of the dental team in mass fatality incidents. Dent Clin North Am 45:271–92. 20. Gahleitner, A., G. Watzek, and H. Imhof. 2003. Dental CT: imaging technique, anatomy, and pathologic conditions of the jaws. Eur Radiol 13:366–76. 21. Abrahams, J. J. 2001. Dental CT imaging: a look at the jaw. Radiology 219:334–45. 22. Fixott, R. H. 2001. The Dental Clinics of North America— Forensic Odontology. W.B. Saunders. 23. Odlum, O. 2001. A method of eliminating streak artifacts from metallic dental restorations in CTs of head and neck cancer patients. Spec Care Dentist 21:72–4. 24. Thali, M. J., T. Markwalder, C. Jackowski, M. Sonnenschein, and R. Dirnhofer. 2006. Dental CT imaging as a screening tool for dental profiling: advantages and limitations. J Forensic Sci 51:113–19. 25. Jackowski, C., A. Lussi, M. Classens, et al. 2006. Extended CT scale overcomes restoration caused streak artifacts for dental identification in CT--3D color encoded automatic discrimination of dental restorations. J Comput Assist Tomogr 30:510–13. 26. Eisenburger, M., J. Hughes, N. X. West, R. P. Shellis, and M. Addy. 2001. The use of ultrasonication to study remineralisation of eroded enamel. Caries Res 35:61–66. 27. Sidler, M., C. Jackowski, R. Dirnhofer, P. Vock, and M. Thali. 2007. Use of multislice computed tomography in disaster victim identification—advantages and limitations. Forensic Sci Int 169:118–28. 28. Thali, M. J., K. Yen, T. Plattner, et al. 2002. Charred body: virtual autopsy with multi-slice computed tomography and magnetic resonance imaging. J Forensic Sci 47:1326–31. 29. Bass, W. M. 1995. Human Osteology: A Laboratory and Field Manual (4th ed.). Columbia: Missouri Archaeological Society. 30. Mall, G., M. Hubig, A. Buttner, J. Kuznik, R. Penning, and M. Graw. 2001. Sex determination and estimation of stature from the long bones of the arm. Forensic Sci Int 117:23–30. 31. De Mendonca, M. C. 2000. Estimation of height from the length of long bones in a Portuguese adult population. Am J Phys Anthropol 112:39–48. 32. Verma, S. S., H. Bharadwaj, T. Zachariah, S. Kishnani, and M. R. Bhatia. 1983. Prediction of body volume by a stepwise linear regression technique. Eur J Appl Physiol Occup Physiol 52:126–30.
D2
Thanatology
CONTENTS D2.1 Decomposition............................................................................................................................................................... 188 D2.1.1 Introduction ..................................................................................................................................................... 188 D2.1.2 Postmortem Imaging Findings in Corpses Demonstrating Advanced Putrefaction and Adipocere Formation ...................................................................................................................................... 188 D2.1.2.1 Skin and Musculature .................................................................................................................... 188 D2.1.2.2 Intracranial Findings ..................................................................................................................... 190 D2.1.2.3 Heart and Great Vessels................................................................................................................. 192 D2.1.2.4 Viscera ........................................................................................................................................... 193 D2.1.3 Conclusions ..................................................................................................................................................... 194 D2.1.4 References ....................................................................................................................................................... 194 D2.2 Investigation of Decomposing Sheep Brain by Means of 1H Magnetic Resonance Spectroscopy: An Attempt toward an Objective PMI Estimation in the Later Postmortem Period .................................................... 194 D2.2.1 Introduction to PMI Estimation ...................................................................................................................... 194 D2.2.1.1 General Aspects of the Time of Death .......................................................................................... 194 D2.2.1.2 State of the Art .............................................................................................................................. 195 D2.2.2 PMI Estimation: Approach by Magnetic Resonance Spectroscopy ............................................................... 196 D2.2.2.1 Magnetic Resonance Spectroscopy in Forensic Sciences ............................................................. 196 D2.2.2.2 Using 1H-MRS of Brain Tissue for PMI Estimation: The Hypothesis .......................................... 196 D2.2.2.3 PMI Estimation by a Single Time-Dependent Variable ................................................................ 197 D2.2.3 Brain MRS ...................................................................................................................................................... 198 D2.2.3.1 Localized Single-Voxel 1H-MRS ................................................................................................... 198 D2.2.3.2 Quantitation of 1H-MR Spectra with LC Model ........................................................................... 198 D2.2.3.3 Application of an Animal Model: Human Brain versus Sheep Brain ........................................... 198 D2.2.4 The Sheep Model at Room Temperature (20nC) ............................................................................................ 200 D2.2.4.1 Experimental Details .................................................................................................................... 200 D2.2.4.2 Sheep Brain Decomposition over Time ......................................................................................... 200 D2.2.4.3 Time Courses of Metabolite Concentrations ................................................................................. 203 D2.2.5 Calculation of PMI.......................................................................................................................................... 203 D2.2.5.1 Combination of Various Metabolites ............................................................................................ 203 D2.2.5.2 Validation of the Sheep Brain Model and the Calculation of PMI................................................ 204 D2.2.6 Discussion and Outlook .................................................................................................................................. 205 D2.2.6.1 Evaluation of the Sheep Brain Model ............................................................................................ 205 D2.2.6.2 Outlook .......................................................................................................................................... 205 D2.2.7 References ....................................................................................................................................................... 206 D2.3 Vital Reactions and Vital Signs .................................................................................................................................... 208 D2.3.1 Introduction ..................................................................................................................................................... 208 D2.3.2 Consciousness and Reaction of a Person toward a Trauma ............................................................................ 208 D2.3.3 Reaction of Tissue toward a Trauma............................................................................................................... 209 D2.3.4 Blood Circulation .............................................................................................................................................210 D2.3.5 Respiration and Ingestion ............................................................................................................................... 212 D2.3.6 Conclusion........................................................................................................................................................216 D2.3.7 References ........................................................................................................................................................216
187 © 2009 by Taylor & Francis Group, LLC
188
The Virtopsy Approach
D2.1 DECOMPOSITION Stephan A. Bolliger and Michael J. Thali
D2.1.1 INTRODUCTION Depending on the season, the environment, and the geographic location, decomposing bodies are a more or less frequent phenomenon. Apart from the obvious distractions such as the offending pungent odor and the sight of the body—which in our experience can pose great difficulties for young and inexperienced (i.e., unaccustomed) forensic pathologists—the decomposition process itself can destroy relevant pathological findings or traces. A badly decomposed corpse is indeed a challenge concerning traditional autopsy techniques [1,2]. The first challenge arises at identification. Whereas visual inspection of certain morphological facial traits or scars of fresh corpses often leads to a suspected identity of the victim due to missing persons registers, which can then be verified using a comparison of the dental or radiological records, fingerprints, or DNA profiling, this decisive first clue as to the deceased’s identity is lacking in putrefied bodies due to the postmortem changes of the face regarding form and color. This problem is especially evident when large numbers of decomposing bodies are to be examined, such as after the Indo-Pacific tsunami catastrophe in December 2004, in which more than 200,000 people from many different countries were killed. In this setting, a computed tomography (CT) scan could have delivered further identification clues such as orthopedic implants or pacemakers rapidly and therefore could have led to an exclusion or inclusion of body identities. By performing this form of screening, the individual caseload composed of different identification procedures such as DNA, fingerprint, and dental identification can be reduced and therefore speed up the identification process. As radiological identification techniques are described in detail in Chapter D1, “Radiologic Identification,” this topic is not further discussed here. Under optimal (i.e., fresh and generally unharmed) conditions, the body’s structures are more or less kept in place by anatomical boundaries. Thus, the forensic examiner can rely on his or her knowledge of the normal human anatomy and clinical pathology skills. In these cases, the form, color, texture, and smell of an organ can give important clues as to an underlying pathology. The decomposing body, depending on the putrefaction state, can render the observation of these basic findings almost impossible. Detecting and discriminating certain suspect odors in a badly decomposed body is, with the exception of a massive intoxication with certain toxins, rarely successful. The form of an organ can also be dramatically altered. Due to putrefaction gases, the object of interest can be turned into a severely bloated, barely recognizable organ of foamy appearance. Moreover, the colonization with insects, especially maggots, can lead to whole organs being eaten away. The individual organ color tends to give way to a universal brown-black discoloration. In advanced putrefaction, the texture of an organ can be completely dissolved, leading to a
© 2009 by Taylor & Francis Group, LLC
liquefaction of entire organ systems. This is especially true for organs rich in lipids, such as the brain. Of the aforementioned problems, the disintegration of organs, the loss of the typical color, and texture and animal infestation cannot be countered. Nevertheless, if preexisting anatomical boundaries such as thin collagenous tissue are respected, the liquefied organs may be assessed to a certain extent. A traditional autopsy destroys these boundaries. This leads to the frequent and frustrating experience of liquefied organs, such as the brain, oozing out of the opened cranial cavity. Obviously, such a liquefied mass can hardly be assessed properly. The main difficulty in the traditional autopsy of cases of advanced putrefaction is therefore the loss of tissue structure, which in turn can lead to a liquefaction of entire organ systems. Postmortem imaging of decomposing bodies has shown promising results [3] in the past. Postmortem multislice CT (MSCT) and magnetic resonance imaging (MRI) leaves the organ boundaries intact. Liquefaction of organs is therefore a minor problem. However, the presence of putrefaction gas can complicate the assessment of CT findings. Generally speaking, while the main problem of traditional autopsies of decomposed bodies is mainly of a fluid nature, in postmortem CT imaging it is of a gaseous nature. The main difficulty of postmortem MRI is the loss of signal in changing chemical properties of the soft tissues. Under damp and oxygen-poor conditions, such as at the bottom of a freshwater body, the decomposition takes a different course. The fatty acids that are released postmortem act as a bactericide to the putrefaction bacteria. Thus, in the absence of bacterial destruction the fat can form an insoluble soap when hydrolyzed from the fat conjugate with bivalent ions such as calcium [4,5]. This gives rise to a wax-like texture of the body fat, also called adipocere. The radiological findings in adipocere are discussed later in this chapter. On the other hand, especially under warm and humid conditions, the decomposition and animal interference eventually leaves clean, blank bones that are devoid of any soft tissues. Depending on the environment in which these bones lay (e.g., a sun-bleached surface, acidic soil), the bones also disintegrate and complete the nutrient cycle. Under conditions with decreased air humidity, such as in cold, windy locations or in heated rooms in winter, corpses can mummify. Obviously, this mummification process can also be man-made as various cultures have demonstrated impressively. The radiological findings of mummified bodies are discussed in Chapter D8.1, “Paleoradiology.”
D2.1.2 POSTMORTEM IMAGING FINDINGS IN CORPSES DEMONSTRATING ADVANCED PUTREFACTION AND ADIPOCERE FORMATION D2.1.2.1 Skin and Musculature Depending on the state of putrefaction, the accumulation of gas in subcutaneous tissues and the muscles is an often encountered phenomenon. This gives rise to a feathery appearance of the subcutaneous fatty tissue and the muscles
Thanatology
189
FIGURE D2.1.3 Macroscopic photo of the subcutaneous fatty tissue of the corpse seen in Figure D2.1.1 and Figure D2.1.2. Note the bubbly aspect of the subcutaneous tissue due to the (postmortem) putrefaction gas emphysema.
FIGURE D2.1.1 Coronal reconstructed MSCT image of the pelvic region of a corpse in a state of advanced putrefaction (postmortem interval roughly 3 weeks at room temperature). Note the feathery appearance of the musculature (green arrows) and foamy subcutaneous fatty tissue (yellow arrows).
in MSCT. Sometimes, minuscule bubbles are seen in the CT images (Figure D2.1.1, Figure D2.1.2, Figure D2.1.3, and Figure D2.1.4). This finding can give the impression of subcutaneous emphysema. A subcutaneous emphysema, which typically consists of an accumulation of air in the soft tissues (i.e., of the skin), is seen in cases where, due to the opening of the venous system, air is embolized into the organism. The
FIGURE D2.1.2 In this axial MR image of the same corpse, the feathery structure of the muscles (green arrows) and the subcutaneous fat (yellow arrows) are barely visible. Note also the accumulation of fluid (in this case putrefaction fluid) in the right thoracic cavity (X).
© 2009 by Taylor & Francis Group, LLC
corresponding MSCT findings have been published previously [6]. The finding of an air embolism is clearly a vital sign and proves that the deceased had lived at the time of the injury infliction. An accumulation of putrefaction gas can appear similarly in CT scans. This is clearly a postmortem finding and is case relevant only in the context of a longer postmortem period. The differentiation between the forensically significant emphysema and the postmortem accumulation of putrefaction gases in the subcutaneous fatty tissue and the muscles can be readily made. First, the pungent odor and the general appearance of the body with regard to decomposition signs permit the assumption of the prevalence of the latter. However, the combination of decomposition gases and a
FIGURE D2.1.4 Macroscopic photo of the subcutaneous fatty tissue of the corpse seen in Figure D2.1.1 and Figure D2.1.2. Note the bubbly aspect of the subcutaneous tissue due to the (postmortem) putrefaction gas emphysema.
190
The Virtopsy Approach
prior (air) emphysema is also possible. The lack of injuries that could have given rise to emphysema deems this possibility rather unlikely. Obviously, one organ system such as the heart should not be assessed in an isolated manner. This rule applies not only to traditional autopsies but also to postmortem imaging. Should signs of an accumulation of gas (air or putrefaction) be seen in the subcutaneous or muscle tissue, then the existence of such gas should be sought for in other, entirely independent organs. In a decomposing body, putrefaction gases are omnipresent as gas microbubbles assembled around fatty cells and muscle fibers. Gas accumulations are also seen inside the body cavities and are, with exception of large amounts of gas in the abdominal cavity, usually discreet. A chemical analysis of the gas can solve persisting uncertainties as the origin of the gas. In corpses undergoing adipocere development, MSCT can demonstrate the distribution of adipocere. This formation induces hyperdensities within the subcutaneous tissue of up to 1,000 Hounsfield units (HU) and higher by deposition of calcium due to saponification of the fatty tissue [7,8] (Figure D2.1.5 and Figure D2.1.6).
FIGURE D2.1.6 MSCT imaging of the body seen in Figure D2.1.5 with a threshold of 250 HU. Note the distribution of the radiological absorbent subcutaneous tissue (red arrows) corresponding to adipocere.
If the postmortem conditions are known, then the postmortem interval can be estimated [9–11]. Decomposing bodies, as mentioned already, often present a dark, brown-black, discolored skin. This makes the external assessment of the presence of hematomas extremely difficult. Postmortem MSCT is an effective method to detect osseous injuries and gas in a body. However, in the visualization of soft-tissue injuries, it is clearly not as effective as an MRI. The main advantage of MRI over MSCT is the detection of soft tissues and their respective injuries. As Yen et al. [12] have already shown, MRI can detect such hematomas of the subcutaneous fatty tissue in fresh corpses. Whether this method is also effective in the assessment of such findings in decomposing bodies is unclear at present and remains subject to further investigations. However, in the evaluation of adipocerous tissue, postmortem MRI proves inferior to postmortem MSCT. The pathognomonic calcifications of adipocerous skin, in addition to putrefaction gases, give rise to a low MRI signal [7] (Figure D2.1.7). This makes the MRI assessment of subcutaneous hemorrhages in adipocere almost impossible and renders this method (at present) ineffective. D2.1.2.2 Intracranial Findings FIGURE D2.1.5 Here, a corpse (postmortem interval in a Swiss lake of almost 3 years) with distinct adipocerous changes is shown. The epidermis is lacking; however, the subcutaneous tissue is present, gray-white, and waxen.
© 2009 by Taylor & Francis Group, LLC
In decomposing bodies, the autoptic assessment of intracranial structures, especially of the brain and the subarachnoid space, is hardly possible. Although the collagenous tissue of
Thanatology
191
FIGURE D2.1.9 This sagittal MSCT image displays intracranial gas accumulations. The green arrows indicate intraparenchymal bubbles, and the yellow arrow indicates the free extracerebral gas. FIGURE D2.1.7 Coronal, T2 weighted (TE-98 ms/TR-4000) MR image of the corpse seen in Figure D2.1.5 and Figure D2.1.6. The decreased signal intensity gives rise to the poor image quality, clearly illustrating the limits of MRI in adipocere development.
the dura mater is rather resistant to putrefactive processes, the brain is, by contrast, highly affected. The opening of the skull gives rise to the loss of the last anatomical boundaries of the liquefied brain, which then oozes out of the cranial cavity and therefore eludes further detailed inspection (Figure D2.1.8). To counter this, a technique was developed in Belgium in which the decomposed body is beheaded [13]. The head is then deep-frozen until solid and then sawed apart in the coronal plane. The two cranial halves are fixed with formalin for later assessment. Today, the technique of the
FIGURE D2.1.8 Here, a brain of a corpse with a postmortem interval of more than 4 weeks is shown. The brain is highly liquefied and leaves the cranial cavity as a papescent mass that eludes further detailed evaluation.
© 2009 by Taylor & Francis Group, LLC
beheading a body for later inspection of the cerebral tissues is generally ethically unacceptable. The no-touch or minimally invasive technique of postmortem imaging can help assess the decomposing brain without unnecessary destruction of the corpse. As the anatomical structures (i.e., skull and cerebral coverings) are left intact, the cerebral tissues tend to keep their anatomical relation to each other, which facilitates the examination immensely. During putrefaction, the overall size of the brain shrinks while putrefaction gases are produced. This gas can accumulate in the subdural and subarachnoid spaces and in the brain itself in the form of bubbles. These bubbles of up to half a centimeter in diameter gave rise to the autopsy term Swiss cheese brain. Indeed, the aspect of tiny bubbles throughout the brain does compare to certain Swiss cheeses. This is easily understandable, as both arise due to microbial gas production. Fortunately, the microbes and the produced gas differ fundamentally (Figure D2.1.9 and D2.1.10). The free intracranial gas in the subdural and subarachnoid space can imitate a pneumocranium. However, if the existence of an
FIGURE D2.1.10 Sagittal MSCT reconstruction highlighting gas further illuminates the findings regarding the intracranial gas seen in Figure D2.1.9. Note the (physiological) air in the airways.
192
FIGURE D2.1.11 This axial MSCT image of a decomposed head displays small bubbles in the cerebral parenchyma (green arrows), giving rise to the term Swiss cheese brain. Free, extracerebral gas (yellow arrow) is seen in the frontal region. The body was found lying on its back. Therefore, the decomposing brain sank to the posterior regions of the intracranial cavity.
open cranial trauma can be excluded, then a pneumocranium can be ruled out (Figure D2.1.11 and D2.1.12). Postmortem MSCT is a useful tool for the evaluation of osseous lesions and for the detection of intracranial gas.
The Virtopsy Approach
FIGURE D2.1.13 T2 weighted (TE-98 ms/TR-4000 ms) MRI clearly demonstrating the superiority of MRI to MSCT with regard to the evaluation of decomposing brains. The cerebral structures are well visible.
MRI, although of limited usefulness in adipocere, is an eminently effective method for the evaluation of cerebral injuries. The anatomical structures are shown astonishingly well [7,8] and suffice to assess macropathological findings such as subarachnoidal, epidural, subdural, and intracranial hemorrhages. These potentially extremely relevant findings can be missed in traditional autopsies (Figure D2.1.13). D2.1.2.3 Heart and Great Vessels As mentioned previously, putrefaction gases are also seen in the great blood vessels and the heart. These give rise to the frequently seen balloon-like appearance of the putrefied heart. The nature of this heart form is quickly clear upon opening the pericardium and the heart, as these structures deflate immediately upon opening. In MSCT, the presence of gas in the right and left heart chambers indicates that this gas—if no preexisting anatomical variation is evident—is of putrefactive nature. This also true for the great arterial and venous blood vessels (Figure D2.1.14, Figure D2.1.15, and Figure D2.1.16).
FIGURE D2.1.12 In this axial MSCT image, a suicidal gunshot injury to the head of a fresh corpse is shown. The overall aspect is similar to Figure D2.1.11 regarding extracerebral gas (yellow arrow)—in this case, air—and dark (air) bubbles (green arrows). The latter are accentuated within the blood vessels, as opposed to the gas bubbles in putrefying brains, where they are omnipresent. Note the entry gunshot wound (red arrow) and the contralateral exit-wound fracture system (orange arrows).
© 2009 by Taylor & Francis Group, LLC
FIGURE D2.1.14 MSCT, axial image showing a cross-section of the heart in a fresh corpse. No gas is seen in the chambers.
Thanatology
193
FIGURE D2.1.15 MSCT, axial reconstruction. Here, a cross-section of the heart of a decomposing body (PMI 3 weeks) is shown. Note the accumulation of gas (arrows) in both right- and left-sided cardiac chambers.
D2.1.2.4 Viscera
FIGURE D2.1.17 This coronal MSCT image shows the abdominal organs in a decomposing body. The organs can be readily located and assessed with regard to injuries. Note the foamy appearance of the visceral organs, especially of the liver (yellow arrow) and the kidneys (green arrows), due to putrefaction gas. The heart (red arrow) is also clearly visible.
Of these, especially the liver displays typical putrefactive changes. The most striking is the foamy appearance, which results from the intraparenchymal production of putrefaction gas similar to the aforementioned Swiss cheese aspect of the decomposing brain. The autoptic extraction of the inner organs can, due to the liquefied state, pose great difficulties. Often, these previously intact organs dissolve when an attempt is made to remove them from the abdomen. This is especially true for the spleen. Postmortem imaging can counter this difficulty. By leaving the abdominal cavity unharmed—as opposed to traditional autopsy techniques— the organs remain more or less intact. MSCT serves to localize these organs and potential foreign bodies. However, this method is only of limited use in the assessment of pathologic changes. MRI, on the other hand, creates a good imaging of the abdominal organs in a similar manner, as seen in the decomposing brain. Using these images, it is possible to determine macroscopically visible changes (Figures D2.1.17 and D2.1.18).
Maggot infestation is, as mentioned before, also a frequently seen phenomenon in rotting corpses. In postmortem MSCT, large accumulations of maggots, as often seen in the airways and in the abdominal cavity, are seen as rather characteristic fluffy regions (Figure D2.1.19).
FIGURE D2.1.16 This axial MSCT demonstrates an air embolism (suicide due to gunshot wound to the head in an upright seated position). The gas—in this case, air—is found only in the right heart chambers (arrow).
FIGURE D2.1.18 Coronal, T2 weighted (TE-98 ms/TR-4000) MR image of the corpse seen in Figure D2.1.19. The internal organs are clearly visible. Note the heart (red arrow), the liver (yellow arrow), and the kidneys (green arrows).
© 2009 by Taylor & Francis Group, LLC
194
The Virtopsy Approach
9. Cotton, G.E., Aufderheide, A.C., and Goldschmidt, V.G. 1987. Preservation of human tissue immersed for five years in fresh water of known temperature. J Forensic Sci 32: 1125–1130. 10. Rothschild, M.A., Schmidt, V., and Pedal, I. 1996. Cadaver lipid: various origins add to the difficulty of assessing postmortem time. Arch Kriminol 197: 165–174. 11. Fiedler, S. and Graw, M. 2003. Decomposition of buried copses, with special reference to the formation of adipocere. Naturwissenschaften 90: 291–300. 12. Yen, K., Vock. P., Tiefenthaler, B., et al. 2004. Virtopsy: forensic traumatology of the subcutaneous fatty tissue; multislice computed tomography (MSCT) and magnetic resonance imaging (MRI) as diagnostic tools. J Forensic Sci 49: 799–806. 13. Knight, B. 1996. Forensic Pathology. New York: Oxford University Press.
FIGURE D2.1.19 The MSCT of a decomposed corpse (PMI over 3 weeks) displays maggot infestation in the region of the carina (yellow arrows). Note the fluffy appearance of the maggot mass.
D2.1.3 CONCLUSIONS Postmortem imaging with MSCT and MRI has several advantages concerning the assessment of decomposing bodies:
D2.2 INVESTIGATION OF DECOMPOSING SHEEP BRAIN BY MEANS OF 1H MAGNETIC RESONANCE SPECTROSCOPY: AN ATTEMPT TOWARD AN OBJECTIVE PMI ESTIMATION IN THE LATER POSTMORTEM PERIOD Michael Ith, Chris Boesch, and Eva Scheurer
D2.2.1 INTRODUCTION TO PMI ESTIMATION Rapid detection of foreign bodies Screening for identification relevant implants Demonstration of gas distribution (as compared with gas embolism) Better visualization of liquefaction-prone organs such as the brain
D2.1.4 REFERENCES 1. Di Maio, V.J.M. 2006. Forensic Pathology. Boca Raton, FL: CRC Press. 2. Spitz, W.U. 1993. Medicolegal Investigation of Death. Springfield, IL: Thomas. 3. Thali, M.J., Yen, K., Schweitzer, W., et al. 2003. Into the decomposed body-forensic digital autopsy using multislicecomputed tomography. Forensic Sci Int 134: 109–114. 4. Takatori, T. 1996. Investigations on the mechanism of human adipocere formation and its relation to other biochemical reactions. Forensic Sci Int 80: 49–61. 5. Takatori, T. 2001. The mechanism of human adipocere formation. Leg Med (Tokyo) 3: 193–204. 6. Jackowski, C., Thali, M., Sonnenschein, M., et al. 2004. Visualization and quantification of air embolism structure by processing postmortem MSCT data. J Forensic Sci 49: 1339–1342. 7. Jackowski, C., Thali, M., Sonnenschein, M., et al. 2005. Adipocere in postmortem imaging using multislice computed tomography (MSCT) and magnetic resonance imaging (MRI). Am J Forensic Med Pathol 26: 360–364. 8. Bohnert, M., Schmailch, A., Faller-Marquardt, M., et al. 1998. Umwandlung des Gehirns und der Gesichtsweichteile in Leichenlipid: morphologische und radiologische Befunde. Rechtsmedizin 8: 135–140.
© 2009 by Taylor & Francis Group, LLC
Scientific investigations on the estimation of the postmortem interval (PMI) have a history going back to the 19th century [1]. Nevertheless, this topic is still a matter of interest and permanent research within the forensic community. Especially in the later postmortem period PMI estimation becomes so complex that to date no objective method has been accepted for use in daily practice. This chapter introduces a new approach toward an objective estimation of PMI with special focus on the later postmortem period. D2.2.1.1 General Aspects of the Time of Death Traditionally, time of death was defined as irreversible cessation of heart and circulatory function. This definition had to be modified for intensive care medicine where circulation and respiration was maintained artificially. As a consequence, intensive care medicine employed a working definition that uses brain death as the moment that characterizes individual death, even when other organs continue to function [2–5]. Nevertheless, cessation of cardiac function still represents the generally accepted time of death and the transition of the body into the so-called supravital state [6]. The postmortem interval is defined as the time span passed since the occurrence of death. Whereas the ascertainment of death is usually performed by a physician, the expertise of a forensic pathologist is required to determine PMI. A first step toward PMI estimation includes determination and documentation of physical and morphological changes due to the cessation of circulation and the onset of autolytic and heterolytic decomposition processes that can be observed
Thanatology
195
as morphological signs on the body. In addition, external parameters such as ambient temperature and humidity, clothing, location, and possibly antemortem diseases that might influence the decomposition process have to be considered. Knowledge of the PMI is essential in criminal investigations for further measures of tracing, collection of evidence, and inclusion or exclusion of possible suspects. In addition, the procedures and results of forensic examinations have a strong impact on relatives and friends of a deceased person [7], influencing the process of mourning and the management of sentiments of real and imagined guilt. Numerous methods have been proposed in the past. However, current methods of PMI estimation—particularly for the later postmortem interval—are far from being straightforward and accurate. Therefore, the boards of prosecution deem a new method extremely important that may solve these problems. On this basis, as well as on the background of increasing willingness to apply new technologies allowing for non- or minimally invasive investigations, magnetic resonance spectroscopy (MRS) represents an ideal complementary method in the toolbox of the forensic pathologist. D2.2.1.2 State of the Art Figure D2.2.1 gives an overview of the methods currently used in daily forensic procedures and shows the period of time during which the individual methods can be applied. The postmortem period can roughly be divided into two phases: (1) the early postmortem interval, when postmortem changes are influenced mainly by the supravital state and autolysis; and (2) the late postmortem interval, when bacterial processes dominate the decomposition process. Section D2.2.1.2.3 discusses some methods particularly aiming at an objective PMI estimation by using chemical analyses.
Entomology Morphology Supravit. signs Cooling Rigidity Lividity 0
10
(Days)
20
FIGURE D2.2.1 The graph represents an overview of forensic methods typically used for PMI estimation. Individual bars represent the approximate periods of time in which the methods can be applied. As far as possible, multiple methods should be applied in parallel in order to increase the accuracy and reliability of the results. The precision of the different methods varies considerably—while the early postmortem interval can be evaluated with reasonable accuracy, estimations for the later period are much less reliable.
© 2009 by Taylor & Francis Group, LLC
D2.2.1.2.1 Early Postmortem Interval Within about 20 to 60 minutes after death, livor mortis is the first reliable sign of death, based on the congestion of red blood cells in the vessels due to gravity [8]. Appearance and possible displacement by manual pressure allow a rough estimate of the PMI during the initial 30 hours [9,10]. Rigidity of the skeletal muscle (i.e., rigor mortis) can be utilized for about 3 days [8,9,11,12]. In addition to the influence of postmortem environmental factors, the cause of death, antemortem factors such as diseases, physical activity before death, and the individual constitution play a significant role on the development of lividity and rigor mortis [13–16]. Therefore, and due to nonuniform examination techniques, lividity and rigor mortis have only limited medicolegal significance [10]. Cooling of the body after death is a very reliable and thoroughly investigated phenomenon that can be used to determine PMIs up to about 30 hours postmortem [17–23]. Considering the weight of the body and environmental factors, the body’s core temperature permits a retro-calculation of a span of time within which death occurred. A recent study proposes a new formula for the estimation of the time of death by measuring the temperature at different localizations of the body and following the model of finite elements. The temperatures of brain and liver are measured noninvasively by means of microwave technique [24,25], and different factors influencing the precision of PMI estimation are investigated [24–26]. Others compare a new averages-based method of short-term PMI estimation with eight other temperature-based methods and find that the former predicts PMI more accurately [27]. The proposal was also made not to use temperature-based methods as a stand-alone tool but to combine these methods with non-temperature-based methods and indicators [28]. However, body temperature approaches ambient temperature approximately 30 hours after death and, thereafter, is no longer of value for assessing time of death. Additionally, supravital reactions can be used for PMI estimation, including the degree to which the musculature can be mechanically and electrically stimulated [29] and the pupils’ reaction to certain pharmacological substances (e.g., acetylcholine) [30,31]. As one of the last observed supravital reactions, the pupil’s reaction can be documented up to a maximum of 46 hours after death. A good overview of the literature on this subject is given by Henssge and Madea [32]. D2.2.1.2.2 Late Postmortem Interval The later postmortem stage, which starts approximately 3 days after death, is determined by different processes leading to a further decomposition of the soft tissues of the body. During that period, either morphological or zoological methods dominate the estimation of PMI. The progress of decomposition is visible in the form of morphological signs on the surface of the body changing and increasing with time postmortem. Despite the fact that the sequence of their appearance has been described frequently [33,34], it is generally accepted that those signs cannot be used as reliable and objective indicators of the PMI [35]. It
196
is recognized that putrefaction and decay strongly depend on external factors (e.g., temperature, humidity) and internal factors (e.g., antemortem treatment and diseases) [36]. In addition, the assessment of morphological signs is far from being objective and strongly depends on the experience of the forensic pathologist [37]. Forensic entomology studies insects that colonize bodies and their environment during the process of decomposition in various waves [38,39]. These waves appear successively in ecological sequences [40], depending on the stage of decomposition of the corpse, and permit conclusions on the PMI [41]. However, the objective definition of general rules for the specific composition of settlement waves is extremely difficult because insect populations vary with region, season, and environment [42–45]. In addition, this method requires enormous effort, including a well-equipped infrastructure to culture the insects and a time-consuming procedure of at least several days before a statement on PMI can be made. While forensic entomologists are essential for the identification of the species, they are extremely rare. Attempts are currently made to identify insects by DNA-profiling [46–48] in order to avoid the expenditure of cultivating the maggots found on the scene. Because of the regional distribution of the species of flies, many research groups concentrate on the analysis of the specific habitat of certain species as to improve the assessment of species to the different settlement waves [39]. D2.2.1.2.3 Attempts to Develop Objective Methods for PMI Estimation: Chemistry of Body Fluids, Different Organs, and Muscle Tissue Immediately after death, the first phase of the body’s natural degradation begins. The collapse of the chemical, physical, and morphologic organization is known as autolysis and results from the action of endogenous enzymes and the cessation of oxygendependent biochemical processes. The failure of the body to maintain homeostasis leads to a breakdown of the internal equilibrium as well as to the degradation of proteins, carbohydrates, and fats. This results in an increase of breakdown products and to the appearance of nonphysiological substances. Chemical analyses for the purpose of PMI estimation based on concentration changes have been performed on body fluids such as blood and blood serum [10,49], cerebrospinal fluid [50,51], and the vitreous humor of the eyes [52,53]. Due to its easy accessibility, several attempts were made to analyze skeletal muscle [54–56]. Above all, the determination of the potassium concentration in the vitreous humor seemed extremely promising due to its independence from ambient temperature and its almost linear rise in concentration up to about 100 hours after death. Unfortunately, metabolic disturbances prior to death can drastically alter this rise in potassium concentration [57]. In particular, the electrolyte content of the vitreous humor increased proportionally with the concentration of blood alcohol [58,59]. Additionally, it has been recognized that the rise in concentration varies not only interindividually but also between the left and right eye in the same individual [60]. As a consequence,
© 2009 by Taylor & Francis Group, LLC
The Virtopsy Approach
various restrictions and precautionary measures had to be formulated [32,53]. New algorithms were recently published [61,62], but the problems concerning the influence of blood alcohol, antemortem metabolic disorders, and the individual variability persist. An attempt has been made to correlate the concentration of electrolytes and other substances in synovial fluid for the determination of PMI [63]. This method generally suffers from the fact that the collection of body fluids can be very difficult depending on PMI and the state in which the individual body is found. Chemical analyses in the late postmortem phase attempted to identify and quantify byproducts of the decomposition of the body. This process is mainly caused by bacteria and includes putrefaction, a reduction process involving the formation of volatile hydrogen compounds, and decay, an oxidative mechanism [34]. As for shorter PMIs, chemical analyses in the later postmortem phase primarily examined body fluids, such as blood, blood serum, and cerebrospinal fluid [64–66], as well as skeletal muscle and liver tissue [67–70]. Some projects concentrated on specific groups of metabolites, such as the decomposition of fatty acids [71,72] or proteins [73–77]. Vass et al. [78] examined various organs, including brain, liver, kidney, heart, and muscle to determine metabolite concentrations of amino acids, neurotransmitters, and byproducts of decomposition. Using a kind of flow chart with decisions based on different metabolite concentrations, the authors concluded that they could estimate PMIs “so accurate that the estimate is limited by the ability to obtain correct temperature data at a crime scene rather than sample variability.” An overview of recent progress of PMI estimation by means of thanatochemistry has been given lately by Madea (page 542) [67].
D2.2.2 PMI ESTIMATION: APPROACH BY MAGNETIC RESONANCE SPECTROSCOPY D2.2.2.1 Magnetic Resonance Spectroscopy in Forensic Sciences Until today only few research articles using MR methods as an investigative tool were related to forensic medicine [79–89]. Three studies investigated relaxation time effects in the early postmortem period, however, and addressed mainly methodological questions of nuclear magnetic resonance (NMR) [85–87]. MRS studies explicitly aimed at postmortem changes were carried out in vitro on brain tissue [79,80], including one using chemical shift imaging methods [83]. The focus of these studies, however, was not the forensic aspect but mainly the influence of physiological changes after death, which then could be transferred to living organisms. D2.2.2.2 Using 1H-MRS of Brain Tissue for PMI Estimation: The Hypothesis For a long time, only a few studies were published that used chemical analyses of brain tissue for PMI estimation
Thanatology
In summary, it can be hypothesized that decomposing brain tissue can be assessed in an objective and noninvasive way using in situ 1H-MRS. The dynamics of postmortem brain metabolite concentration can be followed over a prolonged period of time with repetitive measurements of the same region of interest (ROI) within the same individual. Thus, a sort of calibration curve can be established for each individual metabolite. Finally, the determination of PMIs by means of 1H-MRS will be performed by comparing the metabolite concentrations of a case with unknown time of death with these calibration curves. However, for ethical reasons the decomposition of brain tissue cannot be studied in human bodies over a prolonged period. Therefore, an animal model was used to investigate characteristics and temporal behavior of decomposition. The choice of the animal was mainly influenced by experimental
© 2009 by Taylor & Francis Group, LLC
D2.2.2.3 PMI Estimation by a Single Time-Dependent Variable As shown schematically in Figure D2.2.2, the PMI can be calculated using a single time-dependent variable—for example, the measured concentration of a brain metabolite (y) as a function of time since death (y f(t)). The prerequisite is that the time course can be described by an unequivocal mathematical function. The inverse function t f –1(y) will directly result in the PMI. The function f(t) is determined by performing multiple measurements over time on multiple individual subjects—in our case, sheep heads—under well-controlled conditions. The data of such a series of experiments will allow a statistical analysis with estimation of the confidence limits (represented in Figure D2.2.2 by the dashed lines). By means of the upper and lower confidence limits, the uncertainty of the estimated PMI can be calculated. This principle of using an inverse function can be applied to any measurable variable varying sufficiently with time postmortem. However, several factors influencing the characteristics of the function such as ambient temperature have to be taken into account. The method can additionally be improved
Measured metabolite concentration Calculated PMI
1. Established MR sequences for 1H allow for a robust and easy selection of spectroscopic volumes (voxels) with subsequent absolute quantitation of the metabolites. 2. 1H-MRS provides high relative sensitivity, allowing the selection of small voxels. 3. A majority of clinical MRS investigations are performed on the brain, resulting in a great experimental experience and a large collection of reference data in healthy and pathological brain tissue. 4. To date, about 25 different substances have been characterized in brain tissue and can be observed and quantified by 1H-MRS.
boundary conditions. The brain should be as large as possible, with the head still fitting into the conventional head coil of the MR scanner, and the animal should be easily obtainable in a local slaughterhouse. Finally, the application of different slaughtering techniques to different cattle determined sheep as the animal of choice. Another research group decided to use pig heads, which led to comparable results [96].
(Concentration)
[78,90–93]. Two publications are particularly worth mentioning [78,93], since in both cases multiple products of bacterial metabolism are used to estimate PMI. Daldrup [93] focused on specific amino acids from bacterial metabolism and proposed a formula for the calculation of PMI in a time span of 4 to 20 days postmortem. Despite the limited number of published studies, examination of the brain would be particularly interesting for PMI estimation. On one hand, interindividual differences in tissue composition are very small compared with other tissues of the body, such as skeletal muscle or liver. The variation of interindividual differences is about 5% [94] and can partly be attributed to the aging process [95], which can be corrected for. On the other hand, the skull forms a natural barrier against scavengers or rodents and direct penetration by external microorganism. Even when the brain tissue is in an advanced state of decomposition, the skull guarantees a certain fixation. With regard to the introduction of modern imaging techniques (e.g., MRI, CT, surface scanning) in forensic medicine, it is straightforward to complement these investigations with an equally noninvasive, objective method for PMI estimation. Additional advantages are that MRS not only is carried out with the same hardware equipment but also can even be performed during the same examination. In recent years, clinical application of 1H-MRS proved particularly advantageous to investigate brain tissue:
197
Estimated variation (Time Postmortem)
FIGURE D2.2.2 This diagram shows schematically how PMI is estimated by a single time-dependent variable—that is, a metabolite concentration at a certain time point. The temporal course of this variable (represented by the solid line) is mathematically described by a model function that should be unequivocal. The PMI is calculated using the inverse parameterized function of the temporal model (black arrow). In order to calculate the uncertainty of the PMI (red arrows), the confidence levels (dashed lines) of the model function need to be estimated.
198
by using more than one variable (e.g., metabolites) and by combining the results from the individual PMI estimations. This procedure is explained in detail in Section D2.2.5.
D2.2.3 BRAIN MRS Some basics of MRS are explained in Chapter B2.4.1, “Basics of MRI and Spectroscopy.” D2.2.3.1 Localized Single-Voxel 1H-MRS Heterogeneity of biochemical composition and metabolism of different organs requires an assessment of signals from precisely chosen and confined areas. Therefore, MRS signal acquisition is restricted to a well-defined volume (i.e., voxel or ROI) that can be placed at the desired anatomical location based on a series of scout MR images of this region. From the various methods of volume selection, short echo time 1H-MR single-voxel (SV) spectroscopy was chosen for PMI estimation because it allows for a simultaneous, highly sensitive detection of several metabolites. Excitation of the voxel is performed using a so-called point-resolved spectroscopy (PRESS) sequence [97]. The simultaneously detected water signal of the voxel is used as a tissue-specific internal concentration standard and, thus, allows the calibration of the metabolite signals. This enables the quantification of the metabolites in absolute units, such as millimols per kg wet weight (mmol/kgww) or per volume (mmol/l).
The Virtopsy Approach
each metabolite with a known concentration. LC Model determines the contribution of individual metabolites to a particular spectrum by approximating the real measured spectrum by a linear combination of weighted individual metabolite spectra. The weighting coefficients are then proportional to the measured metabolite concentrations. Obviously, the individual spectra of all metabolites need to be previously determined under identical experimental conditions. Additionally, it is very important for an accurate quantification that all metabolites contributing significantly to the measured spectrum are considered. A detection limit of about 1 mmol/kg can be achieved. The set of metabolites shown in Figure D2.2.3 represents a complete basis set used to fit in vivo brain spectra. D2.2.3.3 Application of an Animal Model: Human Brain versus Sheep Brain As sheep heads served as a model for the decomposing human brain, both must exhibit very similar characteristics concerning tissue composition and decomposition characteristics. Figure D2.2.4 shows a comparison of sheep- and human-brain spectra obtained at two different moments postmortem. Analysis of the spectra reveals contributions from the same metabolites. Major changes compared with in vivo brain spectra are indicated
D2.2.3.2 Quantitation of 1H-MR Spectra with LC Model In most investigations MRS is used not only to identify and characterize certain metabolites but also to determine their relative or—preferably—absolute concentrations, which involves several procedures summarized under the term quantification. A variety of methods exist to quantify MR spectra, all being based on the fact that the area under the resonance line is proportional to the number of nuclei contributing to the respective signal, given appropriate experimental conditions. The most important prerequisite is a fully relaxed MR signal. Thus, if the concentration of nuclei for a particular peak is known, it can be used as a tissue-specific internal concentration standard, as in the example of the water signal. The metabolite concentrations of other substances are then determined by integrating the appropriate resonance lines and calibrating them with the known standard. Unfortunately, overlapping signals often occurring at clinical field strength prevent this simple method. Therefore, the so-called model fitting using prior knowledge constraints is increasingly accepted for accurate quantification in clinical spectroscopy, and several different algorithms have been proposed [98–104]. For PMI estimation, the commercially available software LC Model was used to quantify metabolite concentrations. The basic idea of LC Model is that every single metabolite has its own spectrum that can be acquired separately from
© 2009 by Taylor & Francis Group, LLC
Ace
Gly
Ala
Glol
Asp
GSH
× 0.5
Ch
IMCL
× 0.5
PCh
Lac
× 0.5
GPC
mI × 0.5
Cr
5
sI
PCr
NAA
Eth
NAAG
GABA
PE
Glc
Tau
Glu
Val
Gln 4
3
2
1
4
3
2
1
FIGURE D2.2.3 This set of 25 metabolites is used as a basis set for quantification of in vivo brain spectra with the LC Model, which uses a linear combination of these individual spectra for the fit. The spectra were measured in solutions of known metabolite concentration at a physiological pH and at 37ºC. Alternatively, a set of simulated spectra can be used as well. The measured metabolite concentrations are determined by the weighting coefficients of the linear combination.
Thanatology
199
Sheep
NA
Human Ace Lac NA
Ace Lac
TMA TMA
fTMA
Cr fTMA
4
3
2
1
0
× 0.5
CH × 0.5
CH
Cr
4
3
2
1
0
FIGURE D2.2.4 The left column shows two spectra from sheep brain 2 days (upper spectrum) and 7 days postmortem (lower spectrum). The right column shows two spectra from human brain at corresponding PMIs. Similar metabolic alterations can be observed early as well as later during the decomposition process. After about 48 hours the degradation of NA and the appearance of Ace and Lac are most prominent and can be observed to about the same extent in both, human and sheep brain. After 7 days the absence of signals in the CH region, the disappearance of Cr, and the significant increase of TMA are obvious. At 2.86 ppm the appearance of a singlet that was assigned to free trimethylammonium can be observed in both species. Remarkably, the overall amplitude of the signals shows a roughly fourfold increase.
by arrows. In the early post-mortem phase (upper row) the degradation of N-acetyl-aspartate (NAA) plus N-acetyl-aspartylglutamate (NAAG)—labeled as NA—and the appearance of lactate (Lac) and acetate (Ace) are most obvious and can be observed in both sheep and human brain. Later in the postmortem period, the disappearance of resonances from carbohydrates (CH) and creatine (Cr) plus phosphocreatine (PCr)—labeled as Cr—is found in both species concordantly. Additionally, both spectra show to the same extent an increase of trimethylammonium (TMA) compounds (choline [Ch], phosphocholine [PCh], and glycerophosphocholine [GPC]) and a further increase of Ace as compared with the earlier moment. Most astonishing was the appearance of an initially unknown resonance at 2.8 ppm (free TMA labeled [fTMA]) in sheep and human brain after about 3–4 days postmortem, which still increased with time. The upper spectrum in Figure D2.2.5 was measured in a sheep brain 18 days postmortem. The red trace results from a fit with the LC Model with the basis set of metabolites introduced in Figure D2.2.3. Their difference is represented by the residuum, which is the trace immediately below the measured and fitted spectra. It shows several residual resonances,
© 2009 by Taylor & Francis Group, LLC
4
3
2
1
FIGURE D2.2.5 The upper illustration shows a spectrum acquired from sheep brain 18 days postmortem together with the fit from the LC Model (red trace) and the residuum (difference between original data set and fitted spectrum) just below. The basis set of metabolites as given in Figure D2.2.3 was used to fit the spectrum. The arrows indicate several resonances that were not fitted, indicating that one or more metabolites lack in the used basis set. The lower illustration shows the identical spectrum, however, with a basis set complemented by the metabolites introduced in Figure D2.2.6. Since the residuum shows no apparent residual resonances, the use of this extended basis set seems to be appropriate.
which clearly indicates that an incomplete basis set of metabolites was used, and that the missing substances have to be identified. Since at this postmortem period signs of bacterial decomposition are usually observed on the body, bacterial metabolism presumably also affects brain-tissue decomposition. That this was the reason of the newly appeared metabolites was additionally supported by the facts that they do not appear in normal or pathologically altered in vivo spectra and that they can be detected in human as well as in sheep brain. Thus, high-resolution NMR experiments were performed on sheep brain extracts for substance identification, which revealed the presence of butyrate (But), isobutyrate (Ibut), fTMA, propionate (Prop), and succinate (Suc) in addition to the metabolites of the basis set. Details about those experiments are presented in Ith et al. [102]. The metabolite patterns of the five newly identified metabolites measured in aqueous solution are presented in Figure D2.2.6. Inclusion of these metabolites into the basis set leads to a significant improvement of the fit quality, which can be seen in the sheep brain spectrum in the lower part of Figure D2.2.5. All of those metabolites were also found in decomposing human brain. In conclusion, as the residuum showed no remaining unfitted resonances in human brain, no additional yet undetected metabolites significantly contribute to postmortem spectra of human brain. The extended basis
200
The Virtopsy Approach
× 0.5
fTMA
× 0.5
Suc But Prop
×4 4
Ibut 3
2
1
FIGURE D2.2.6 Metabolite patterns of the five metabolites, appearing typically in the later postmortem period when bacterial decomposition has started within the brain tissue, are presented. Whereas Suc can also be observed in vivo (e.g., in urine), fTMA, But, Prop, and Ibut were exclusively observed in postmortem spectra.
set (consisting of the metabolites from the conventional basis set supplemented with the five additional metabolites) turned out to be adequate for the quantification of metabolites in all PMI experiments—in the early as well as in the later postmortem period.
D2.2.4 THE SHEEP MODEL AT ROOM TEMPERATURE (20nC) D2.2.4.1 Experimental Details Eight sheep heads were obtained during regular slaughtering in a local slaughterhouse. To prevent loss of cerebrospinal fluid and to avoid direct bacterial contamination of the brain, the spinal canal was sealed by plasticine immediately after decapitation. This precaution turned out to be very important as a tiny amount of air entering the spinal canal and the ventricles made it impossible to acquire spectra of good quality (see Section D2.2.4.2). Each head was placed in a plastic container and stored at room temperature (21 o 3nC) for the entire period of investigation of up to 18 days postmortem. Single-voxel 1H-MRS was performed on a 1.5 T wholebody MR system (GE SIGNA, Milwaukee, WI) using a standard quadrature head coil. Spectra were obtained with a short echo time PRESS sequence (repetition time TR 3 s, echo time TE 20 ms) with water and additional outer volume suppression. The ROIs were placed in the frontal lobe (10 r 10 r 10 mm3) and the parieto-occipital region (10 r 10 r 20 mm3) of the sheep brain, as shown in Figure D2.2.7. This yields a balanced mix of gray matter (GM) and white matter (WM) of GM/WM y 1 for both locations. All results shown in this chapter are from the parieto-occipital voxel due to better spectral quality. Following eddy current correction, spectra were fitted with the LC Model using an extended basis set of metabolites
© 2009 by Taylor & Francis Group, LLC
FIGURE D2.2.7 Typical size and positioning of two exemplary voxels are shown in this cross-section through a sheep brain. The inlet in the upper right corner demonstrates the slice position in a sagittal reference image. Images were recorded using a T2-weighted fast spin echo sequence (TR = 6000 ms, TE = 105 ms). Spectra in the frontal lobe were recorded from a 10 × 10 × 10 mm3 ROI, in the parieto-occipital region from ROIs of up to 10 × 10 × 20 mm3.
according to the description in Section 2.2.3. The spectra were quantified using the fully relaxed water signal as the internal concentration standard, correcting for contributions from cerebrospinal fluid spaces [94] and taking into account the GM/WM ratio determining the water content in the voxel. Water content and relaxation times of metabolite signals were assumed to be constant throughout the entire period of investigation. Even though this assumption might not be true during decomposition when tissue structures are destroyed, water is still the best internal reference. Although metabolite contents could be expressed in molar units (i.e., mmol/kg solvent), which would be less susceptible to tissue degradation, the ease of comparability with in vivo concentrations would be lost. D2.2.4.2 Sheep Brain Decomposition over Time Figure D2.2.8 demonstrates a typical time course of sheep brain decomposition. The time points are chosen to show the typical temporal development of tissue degradation and the corresponding spectral changes but also critical moments when the acquisition of good spectra was difficult due to the formation of gas bubbles. On the left side of Figure D2.2.8 a series of T2-weighted images (liquid appears bright), together with the location of the ROIs, is shown. Tissue degradation can clearly be observed as blurring and loss of texture in the brain. Swelling of tissue structures immediately postmortem leads to the extrusion of liquid from the cerebral furrows (compare the pictures of day 1 and day 2 postmortem). Further
Thanatology
201
Intensity (a.u.)
Day 1 400 300 200 100 0 4
3
2
0
Day 2
500 Intensity (a.u.)
1
400 300 200 100 0 4
3
2
1
0
Intensity (a.u.)
Day 3 500 400 300 200 100 0 4
3
2
1 0 Day 6
4
3
2
1
Intensity (a.u.)
1000 800 600 400 200
Intensity (a.u.)
0 0
Day 18
2500 2000 1500 1000 500 0 4
3 2 1 Frequency (ppm)
0
FIGURE D2.2.8 A time series of images and corresponding spectra from an individual sheep head shows the development of brain tissue decomposition up to 18 days postmortem. The white box indicates the ROI where the spectra originate from. Due to formation of gas bubbles within the brain, the ROI had to be reduced in size and shifted to the contralateral side between the measurements at 2 and 3 days postmortem, respectively.
decomposition leads to an accumulation of liquid within the skull (images from day 6 and day 18 postmortem). The aim was to keep the ROI at the same anatomical position throughout the entire study. However, due to the formation of gas bubbles owing to bacterial decomposition, the voxel had to be shifted or reduced in size or both. When gas bubbles appeared within a ROI, first voxel size was reduced. If this measure was not sufficient, the position of the voxel was shifted, and as a last measure, such as to maintain the GM/ WM ratio within the ROI, the voxel was displaced to the contralateral side as shown in the example at day 3 postmortem.
© 2009 by Taylor & Francis Group, LLC
After such a displacement or size reduction the voxel position and size were kept constant for the rest of the study. The formation of gas bubbles was regularly observed after about 3 days postmortem, coinciding with the supposed onset of bacterial metabolism. In some cases acquisition of spectra was impossible for several days and would have resulted in greatly enlarged line widths. In the spectrum acquired at 6 days postmortem the line width is still quite large, leading to inaccurate quantification and results. As a consequence, fewer measurements were performed in the period between about 80 and 150 hours postmortem, as seen in Figure D2.2.9.
202
The Virtopsy Approach
200
Ace
Ala
20
150
15
100
10
50
5 0
0 0 80
100
200
300
0
400
But
20
60
15
40
10
20
5
0
50
100
150
GABA
0 100
10
200
0
400
300
4
GSH
8
20
40
60
80
100
200
300
400
Ibut
3
6 2 4 1
2 0
0 50
0
100
150
0
NAA & NAAG 6
60
fTMA
50 40
4 30 20
2
10 0
0 0
40
20
40
60
80
60 15
Prop
80
100
120
140
160
180
Val
30 10 20 5 10 0
0 100
200
300
400
0
50
100
150
200
250
FIGURE D2.2.9 The concentrations of 10 selected metabolites are shown as a function of the time elapsed since death. The data originate from all eight sheep heads, and each individual measurement is represented by a circle. Analytical functions were fitted to the data within the shown time frame according to Table D2.2.1 and represent the time courses of the selected metabolites (solid line). Confidence intervals of the fits are indicated by the dashed lines. The fits were restricted to time periods where concentration changes showed unequivocal behavior and scattering of the data points was limited. For Ace, But, GSH, Ibut, and Prop all data points were used for the determination of analytical functions.
© 2009 by Taylor & Francis Group, LLC
Thanatology
203
After this period, signal intensity of the spectra generally increased significantly, which is obvious when comparing the scaling of the y-axis. D2.2.4.3 Time Courses of Metabolite Concentrations Evaluation of the spectra revealed reproducible time courses for some metabolite concentrations. Figure D2.2.9 presents metabolites that show an unequivocal, continuous, temporal behavior within a certain time period, making them suitable for PMI estimation according to the criteria formulated in Section D2.2.5.1 Please note that only Ace, But, Ibut, and Prop can be used throughout the entire observed postmortem period—that is, up to about 400 hours postmortem. The other metabolites showed either unequivocal behavior, such as a drop in concentration, or had no reproducible time courses with a strong scattering of the data after a certain time. For analytical PMI estimation, parameterized model functions were fitted to the data of the 10 metabolites shown in Figure D2.2.9. Model functions were kept as simple as possible (linear, quadratic, exponential, or logistic) with a minimal number of parameters leading to an appropriate description of the concentration changes. Biochemical processes were not considered for choosing a particular function. The functions are summarized in Table D2.2.1 for all 10 metabolites. Limits for concentrations and times were applied to restrict the fit to the meaningful parts of the measured time courses. In addition, upper and lower confidence limits for the model functions were calculated (represented by the dashed lines in Figure D2.2.9).
D2.2.5 CALCULATION OF PMI D2.2.5.1 Combination of Various Metabolites While PMI calculation based on a single measured parameter (e.g., metabolite concentration) was described in Section D2.2.2.3, this section presents a strategy to combine PMI calculations from multiple parameters. Keeping in mind the theoretical scheme for PMI estimation as shown in Figure D2.2.2, it is obvious that its application to the different time courses and confidence limits as shown in Figure D2.2.9 will lead to different PMIs (Ti, or the calculated PMI from an individual metabolite) as well as to different variations (Vari or the PMI variation of an individual metabolite) for each metabolite. The simplest way to combine the PMIs from the different metabolites (Ti) would be to calculate an average and a standard deviation (SD). Thereby the different accuracies of the individual PMI calculations (Vari)—that is, various confidence limits due to the different scattering of the data—would be neglected. A different approach for the calculation of the final PMI (Testimated, or calculated PMI from a combination of metabolites) with respect to Vari is proposed in Equation 1. T
Testimated
3 Vari
i
i
1 3 Var i
The formula for Testimated still represents an average of the different Ti; however, the individual contributions to Testimated (Ti)
tAble d2.2.1 Concentration limits for fit
138 1 exp( (t 153) / 35.1)
Time limits for fit [h]
ACE < 70
-
f (t )
3.27 0.000435 t 2
ALA < 20
t < 200
f (t )
2.56 0.00146 t
-
t < 100
-
t < 200
-
t < 300
TME < 30
50 < t < 200
BUT < 70
t > 50
-
-
Variable
Transformation
Acetate (ACE)
-
f (t )
Alanine (ALA)
-
-Aminobutyrate (GABA) Glutathione (GSH) Valine (VAL) Free Trimethyl ammonium (TME) Butyrate (BUT) Isobutyrate (IBUT) Propionate (PROP) N-acetyl-aspartate + N-acetyl-aspartylglutamate (NAA+NAAG)
-
Model
f (t )
-
f (t )
-
f (t )
-
f (t )
IBUT 0.5
f (t )
-
f (t )
-
f (t )
2
0.212 0.0320 t
0.583 0.000148 t 2 7.25 0.205 t
0.119 0.000435 t
2
0.841 0.0000144 t 1.93 0.0851 t
exp(1.86 0.0314 t )
2
PROP < 30
t > 20
NAA+NAAG > 2.5
t < 100
Parameterized model functions were used to describe the time courses of the ten metabolites shown in Figure D2.2.8. As biochemical mechanisms are not yet understood, functions with the smallest number of parameters that still lead to an appropriate description of the time course were chosen.
© 2009 by Taylor & Francis Group, LLC
(1)
i
204
The Virtopsy Approach
Average Variance (h2) 1000
800
600
400
200
0
are inversely weighted by their variation Vari. Hence, Ti from metabolites with small variations (e.g., from metabolites with reproducible time courses and low data scattering) contribute more to Testimated than Ti with a large variance Vari. The final variance of the weighted prediction (Var(Testimated)) and its SD (i.e. square root of Var(Testimated)) were calculated according to Equation 2.
Ace, Ala, But, fTMA, Prop Ala, But, fTMA, Prop Ace, But, fTMA, Prop Ace, Ala, But, Prop Ace, Ala, fTMA, Prop
Var (Testimated )
1 1 3 Var i
Ace, Ala, But, fTMA
(2)
But, fTMA, Prop
i
Ala, But, Prop Ala, fTMA, Prop
Statistical analysis of all 10 metabolites shown in Figure D2.2.9 and parameterized according to Table D2.2.1 revealed that a combination of Ace, Ala, But, fTMA, and Prop leads to the best results. Combinations of more or different metabolites did not improve Testimated or Var(Testimated).
Ala, But, fTMA Ace, But, Prop Ace, fTMA, Prop Ace, But, fTMA Ace, Ala, Prop
D2.2.5.2 Validation of the Sheep Brain Model and the Calculation of PMI
Ace, Ala, fTMA Ace, Ala Ace, fTMA Ace, But Ace, Prop Ala, fTMA Ala, But Ala, Prop But, fTMA fTMA, Prop But, Prop 1
0.8
0.6
0.4
0.2
0
In order to validate the calculation of PMI, Testimated (y-axis) was compared with the true PMIs (x-axis) for each and every measurement in the sheep model, illustrated in Figure D2.2.10. The aim was to demonstrate that the predicted PMIs using Equation 1 for a combination of selected metabolites (Ace, Ala, But, fTMA, and Prop) are consistent with the data used to generate the model functions. Whereas below 250 hours the calculated PMIs coincide with the line representing unity (red line), the prediction deviates to lower values than their corresponding true values above 250 hours postmortem. The correlation coefficient is r 0.93 for the entire investigated period (0–300 h postmortem) and slightly improves for data below 250 hours (r 0.97).
Ace, Ala, But
Correlation Coefficient
FIGURE D2.2.11 This graph displays the correlation coefficients (in light blue; axis on the bottom) and the average variance (in dark blue; axis on the top) of the correlation between true PMI and predicted PMI for various combinations of metabolites. The bars corresponding to the metabolite combination of Ace, Ala, But, fTMA, and Prop represent the same data as in Figure D2.2.10. The fact that the correlation coefficient remains constant even when only two metabolites are used demonstrates the robustness of the method.
(Hours)
200
100
0 0
100
200 (Hours)
FIGURE D2.2.10 This graph shows the correlation of true PMI (x-axis) versus predicted PMI (y-axis) based on 1H-MRS in the sheep-head model. The calculation was based on the combination of five metabolites: Ace, Ala, But, fTMA, and Prop. Predicted PMIs are shown with a confidence interval of o 2 SD. The line of identity (predicted PMI is equal to the true PMI) is shown in red. A correlation of r = 0.97 was found for times below 250 hours. For PMIs > 250 hours the prediction clearly deviates from the identity line, showing that predicted values underestimate the true PMI.
© 2009 by Taylor & Francis Group, LLC
The systematic underestimation of true PMIs above 250 hours postmortem may occur for several reasons (discussed in detail by Scheurer et al. [103]), including that the time courses of most of the used metabolite concentrations reach a plateau, making Testimated very susceptible to small variations of the measured concentration. Evidently, Vari also increases with flattening of the concentration curve. Another reason is that some metabolites show equivocal time courses; that is, the concentration decreases after a certain time. To determine the model functions, the equivocal parts were omitted according to the restrictions given in Table D2.2.1. However, for a given measurement it cannot be decided whether an individual metabolite concentration was measured with increasing or decreasing concentration. Consequently, the
Thanatology
contribution of such metabolites to Testimated leads to a systematic underestimation of true PMI. As mentioned previously, an inclusion of more than five or a different choice of metabolites would not improve the accuracy of PMI estimation. Figure D2.2.11 demonstrates the effect of reducing the number of metabolites on the quality of the estimated PMI. Surprisingly, correlation and average variance of true versus estimated PMI turned out to be very robust. Even the inclusion of only two metabolites still leads to a correlation coefficient of r 0.87 (in the case of the combination of Ace and Prop), which is comparable to the one obtained with five metabolites. While the correlation coefficients are almost stable, the average variance increases when less metabolites are included. However, the model and the method of PMI estimation by combination of several metabolites show a convincing robustness.
D2.2.6 DISCUSSION AND OUTLOOK D2.2.6.1 Evaluation of the Sheep Brain Model Every new technology or method has to demonstrate that the new approach represents an improvement or complement to current methods. A recent publication from Henssge and Madea [104] specifically focuses on the basic principles of PMI estimation with regard to applicability in daily forensic casework. They clearly formulate claims and conditions for a methodology as well as criteria that have to be fulfilled if a method shall gain practical relevance. The main criteria are as follows: 1. The measurements have to be objective and quantitative. 2. PMI estimation has to based on a mathematical description that allows influencing factors to be considered quantitatively. 3. The precision of PMI estimation has to be quantified and proven by statistical analysis. 4. The method has to be validated on independent case material. PMI estimation by 1H-MRS shall be discussed with respect to these criteria. The objectivity of MRI and MRS has already been shown thoroughly in the literature, including this book. While the measurement of an MR spectrum simply consists of the objective acquisition of raw data, the quantitation of the spectra includes necessarily some subjective decisions. This includes the choice of the quantitation algorithm, which in the presented work is the LC Model, and related aspects, such as the selection of a basis set of metabolites. In addition, the choice of a particular model function describing the time course of a metabolite (Figure D2.2.9 and Table D2.2.1) as well as the combination of the individual PMIs requires some subjective decisions. However, since the raw spectra are observer independent and the choice of the model functions is done once, the method has an unequaled objectivity.
© 2009 by Taylor & Francis Group, LLC
205
The robustness of the presented method to estimate PMI was shown yielding accurate results at least up to 250 hours postmortem. The variation of the estimated PMI easily surpasses conventional methods applied in the later postmortem period—for example, the interpretation of putrefaction signs in combination with criminological information. Obviously, the applicability of the method to human bodies is most important. Initial results demonstrated that the methodology of PMI estimation by 1H-MRS can be used for human cases [102,103]; however, an application to a larger cohort of cases is necessary. The main problem concerning validation of the model is a missing gold standard for PMI in the later postmortem interval. Therefore, the estimated PMIs will always have to be compared with “true” PMIs with extremely large variations. To date, postmortem decomposition of brain tissue has been investigated at room temperature only. Therefore, the model functions for metabolite behavior can only be applied to human cases with comparable ambient conditions. Since all processes leading to brain-tissue degradation are temperature dependent, all model functions y(t) that describe the temporal behavior of metabolite concentrations are actually functions of both time and temperature y(t,T). However, the sheep brain model is currently being investigated at different temperatures in order to determine the appropriate model functions. Investigations at low temperatures might prove particularly interesting and valuable. On one hand, all processes are expected to proceed at a much slower rate, which probably allows a thorough investigation or even separation of the individual processes (e.g., biochemical pathways, bacterial metabolism). On the other hand, the presented method might also prove valuable in the earlier postmortem period when—at low ambient temperatures the body’s core temperature approximates ambient temperature quickly, and the interval during which a temperature-based method can be applied is very short. Exactly in these cases postmortem biochemistry slows down, thus extending the time span when accurate PMI estimation by MRS is possible. As an example, the reproducible decay of NAA NAAG over time can be used for PMI estimation at room temperature only up to about 80 hours postmortem. When at low temperatures this decrease runs significantly slower, accurate PMI estimation is enabled for much longer periods. As a conclusion, the proposed method represents not only an objective PMI estimation in the later postmortem period but under certain circumstances also an ideal complement to established methods in the early postmortem period. D2.2.6.2 Outlook Conventional MRI for radiological diagnosis is nowadays routinely used in medical institutions such that an introduction into forensic medicine is straightforward. In contrast, PMI estimation by means of MRS currently seems to be of scientific interest only, as both the methodology of MRS and the infrastructure are not integrated in forensic institutes.
206
However, 1H-MR spectra can theoretically be acquired on all clinical MR scanners, and the principle and the application of the method of PMI estimation itself are simple. Widespread application of MRS for forensic purposes would be possible if knowledge and experience on this method would increase and if collaboration with MR experts would be fostered.
D2.2.7 REFERENCES 1. Casper, J.L., Zeit des Todes, P., in Handbuch der gerichtlichen Medizin, 8th ed., Liman Carl, Ed., Verlag von August Hirschwald, Berlin, 1889, 14. 2. Capron, A.M. Brain death—well settled yet still unresolved, N. Engl. J. Med., 344, 1244, 2001. 3. Spann, W. and Liebhardt, E., [Resuscitation and determination of the time of death], Munch. Med. Wochenschr., 108, 1410, 1966. 4. Kurthen, M., Linke, D.B., and Reuter, B.M., [Brain death, death of the cerebral cortex or personal death? On the current discussion of brain-oriented determination of death], Med. Klin., 84, 483, 1989. 5. Dirnhofer, R. Ist der Hirntod wirklich Tod? Rechtsmedizinische Gedanken zur Pathophysiologie des Sterbens, Schweiz. Aerztezeitschrift, 1295, 1997. 6. Madea, B. and Henssge, C., Supravitalität, Rechtsmed., 1, 117, 1991. 7. Plattner, T., Scheurer, E., and Zollinger, U., The response of relatives to medicolegal investigations and forensic autopsy, Am. J. Forensic Med Pathol., 23, 345, 2002. 8. Mallach, H.J., Zur Frage der Todeszeitbestimmung, Berl. Med., 18, 577, 1964. 9. Mallach, H.J. and Mittmeyer, H.J., [Rigor mortis and livores. Estimation of time of death by use of computerized data processing], Z. Rechtsmed., 69, 70, 1971. 10. Schleyer, F., Leichenveränderungen. Todeszeitbestimmung im früh-postmortalen Intervall, in Gerichtliche Medizin, 2d ed., Springer Verlag, Berlin, 1975, 55. 11. Von Hofmann, E. Die forensisch wichtigsten Leichenerscheinungen, Vijschr. Gerichtl. Med., 25, 229, 1876. 12. Krompecher, T. and Fryc, O., [Determination of the time of death based on rigor mortis], Beitr. Gerichtl. Med., 37, 285, 1979. 13. Krompecher, T., et al., Experimental evaluation of rigor mortis. VI. Effect of various causes of death on the evolution of rigor mortis, Forensic Sci. Int., 22, 1, 1983. 14. Kussmaul, A., Über die Totenstarre und die ihr nahe verwandten Zustände von Muskelstarre, mit besonderer Berücksichtigung auf die Staatsarzneikunde., Prakt. Heilk., 13, 67, 1856. 15. Forster, B., et al., Tierexperimente und an menschlichen Leichen gewonnene Daten zur Frage der Totenstarre, Krim. forens. Wissen., 13, 35, 1974. 16. Fechner, G., Koops, E., and Henssge, C., [Cessation of livor in defined pressure conditions], Z. Rechtsmed., 93, 283, 1984. 17. Seydeler, R., Nekrothermometrie, Prag Vijschr., 104B, 137, 1869. 18. Rainy, H., On the cooling of dead bodies as indicating the length of time that has elapsed since death, Glasgow Med. J., 1, 323, 1869. 19. Marshall, T.K. and Hoare, F.D., Estimating the time of death: The rectal cooling after death and its mathematical expression, J. Forensic Sci., 7, 56, 1962. 20. Shapiro, H.A., The post-mortem temperature plateau, J. Forensic Med., 12, 137, 1965.
© 2009 by Taylor & Francis Group, LLC
The Virtopsy Approach
21. Henssge, C. [Precision of estimating the time of death by mathematical expression of rectal body cooling], Z. Rechtsmed., 83, 49, 1979. 22. Henssge, C., et al., Experiences with a compound method for estimating the time since death. I. Rectal temperature nomogram for time since death, Int. J. Legal Med., 113, 303, 2000. 23. Marty, W., Thermographie und Thermometrie in der Forensik mit besonderer Berücksichtigung der Todeszeitbestimmung, Habil.Schrift, Zürich, 1995. 24. Al-Alousi, L.M. et al., Multiple-probe thermography for estimating the postmortem interval: I. Continuous monitoring and data analysis of brain, liver, rectal and environmental temperatures in 117 forensic cases, J. Forensic Sci., 46, 317, 2001. 25. Al-Alousi, L.M. et al., Multiple-probe thermography for estimating the postmortem interval: II. Practical versions of the Triple-Exponential Formulae (TEF) for estimating the time of death in the field, J. Forensic Sci., 46, 323, 2001. 26. Al Alousi, L.M. et al., Factors influencing the precision of estimating the postmortem interval using the triple-exponential formulae (TEF). Part I. A study of the effect of body variables and covering of the torso on the postmortem brain, liver and rectal cooling rates in 117 forensic cases, Forensic Sci. Int., 125, 223, 2002. 27. Nelson, E.L., Estimation of short-term postmortem interval utilizing core body temperature: a new algorithm, Forensic Sci. Int., 109, 31, 2000. 28. Henssge, C., et al., Experiences with a compound method for estimating the time since death. II. Integration of non-temperature-based methods, Int. J. Legal Med., 113, 320, 2000. 29. Madea, B. and Henssge, C., Electrical excitability of skeletal muscle postmortem in casework, Forensic Sci. Int., 47, 207, 1990. 30. Prokop, O. and Fünfhausen, G., Ueber die postmortale Pupillenreaktion auf pharmakologische Reize, Beitr. Modern. Therap., 2, 469, 1960. 31. Klein, A. and Klein, S., Die Todeszeitbestimmung am menschlichen Auge, Med.Diss., Dresden, 1978. 32. Henssge, C. and Madea, B., Methoden zur Bestimmung der Todeszeit an Leichen, Schmidt-Römhild Verlag, Lübeck, 1988. 33. Strassmann G., Über Leichenveränderungen, autolytische, Fäulnis- und Verwesungsvorgänge, Dtsch Z ges gerichtl Med, 3, 359, 1924. 34. Berg, S., Leichenzersetzung und Leichenzerstörung, in Gerichtliche Medizin, 2d ed., Springer Verlag, Berlin, 1975, 62. 35. Schneider, V. and Riese, R., Fäulnisveränderungen an Leichen. Ein Beitrag zur Todeszeitbestimmung., Kriminalistik, 34, 297, 1980. 36. Mann, R.W., Bass, W.M., and Meadows, L., Time since death and decomposition of the human body: variables and observations in case and experimental field studies, J. Forensic Sci., 35, 103, 1990. 37. Burton, J.F., Fallacies in the signs of death, J. Forensic Sci., 19, 529, 1974. 38. Mégnin, P., La faune des cadavres: Application de l’entomologie à la médicine légale, Masson, Paris, 1894. 39. Arnaldos, I., et al., An initial study on the succession of sarcosaprophagous Diptera (Insecta) on carrion in the southeastern Iberian peninsula, Int. J. Legal Med., 114, 156, 2001. 40. Smith, K.G.V., A Manual of Forensic Entomology, British Museum, Natural History, London, and Cornell University Press, Ithaca, NY, 1986.
Thanatology
41. Benecke, M., [Expert insect identification in cases of decomposed bodies], Arch. Kriminol., 198, 99, 1996. 42. Grassberger, M. and Reiter, C., Effect of temperature on Lucilia sericata (Diptera: Calliphoridae) development with special reference to the isomeg, Forensic Sci. Int., 120, 32, 2001. 43. Campobasso, C.P., Di Vella, G., and Introna, F., Factors affecting decomposition and Diptera colonization, Forensic Sci. Int., 120, 18, 2001. 44. Byrd, J.H. and Allen, J.C., The development of the black blow fly, Phormia regina (Meigen), Forensic Sci. Int., 120, 79, 2001. 45. Goff, M.L., Brown, W.A., and Omori, A.I., Preliminary observations of the effect of methamphetamine in decomposing tissues on the development rate of Parasarcophaga ruficornis (Diptera: Sarcophagidae) and implications of this effect on the estimations of postmortem intervals, J. Forensic Sci., 37, 867, 1992. 46. Malgorn, Y. and Coquoz, R., DNA typing for identification of some species of Calliphoridae. An interest in forensic entomology, Forensic Sci. Int., 102, 111, 1999. 47. Vincent, S., Vian, J.M., and Carlotti, M.P., Partial sequencing of the cytochrome oxydase b subunit gene I: a tool for the identification of European species of blow flies for postmortem interval estimation, J. Forensic Sci., 45, 820, 2000. 48. Wells, J.D., Pape, T., and Sperling, F.A., DNA-based identification and molecular systematics of forensically important Sarcophagidae (Diptera), J. Forensic Sci., 46, 1098, 2001. 49. Coe, J.I., Postmortem chemistries on blood with particular reference to urea nitrogen, electrolytes, and bilirubin, J. Forensic Sci., 19, 33, 1974. 50. Coe, J.I., Postmortem chemistry: practical considerations and a review of the literature, J. Forensic Sci., 19, 13, 1974. 51. Endo, T., et al., Postmortem changes in the levels of monoamine metabolites in human cerebrospinal fluid, Forensic Sci. Int., 44, 61, 1990. 52. Madea, B., Henssge, C., and Staak, M., [Postmortem increase in potassium in the vitreous humor. Which parameters are suitable as indicators of antemortem agonal electrolyte imbalance?], Z. Rechtsmed., 97, 259, 1986. 53. Coe, J.I. Vitreous potassium as a measure of the postmortem interval: an historical review and critical evaluation, Forensic Sci. Int., 42, 201, 1989. 54. Mayer, M. and Neufeld, B., Post-mortem changes in skeletal muscle protease and creatine phosphokinase activity—a possible marker for determination of time of death, Forensic Sci. Int., 15, 197, 1980. 55. Mittmeyer, H.J., [Determination of the myo-albumin content. A possibility to determine the hour of death], Z. Rechtsmed., 84, 233, 1980. 56. Mittmeyer, H.J., [Muscle electrophoretic study in the determination of the time of death], Beitr. Gerichtl. Med., 38, 177, 1980. 57. Sturner, W.Q., The vitreous humor, postmortem potassium changes, Lancet, 807, 1963. 58. Sturner, W.Q., et al., Osmolality and other chemical determinations in postmortem human vitreous humor, J. Forensic Sci., 17, 387, 1972. 59. Devgun, M.S. and Dunbar, J.A., Biochemical investigation of vitreous: applications in forensic medicine, especially in relation to alcohol, Forensic Sci. Int., 31, 27, 1986.
© 2009 by Taylor & Francis Group, LLC
207
60. Balasooriya, B.A., St Hill, C.A., and Williams, A.R., The biochemistry of vitreous humour. A comparative study of the potassium, sodium and urate concentrations in the eyes at identical time intervals after death, Forensic Sci. Int., 26, 85, 1984. 61. Munoz, J.I., et al., A new perspective in the estimation of postmortem interval (PMI) based on vitreous, J. Forensic Sci., 46, 209, 2001. 62. Tagliaro, F., et al., Potassium concentration differences in the vitreous humour from the two eyes revisited by microanalysis with capillary electrophoresis, J. Chromatogr. A, 924, 493, 2001. 63. Madea, B., Kreuser, C., and Banaschak, S., Postmortem biochemical examination of synovial fluid—a preliminary study, Forensic Sci. Int., 118, 29, 2001. 64. Mittmeyer, H.J. and Schwend, J., [Time-dependent changes in the protein spectrum of blood from living and dead persons by means of electrophoresis in cellulose acetate], Z. Rechtsmed., 81, 19, 1978. 65. Schleyer, F. and Pioch, W., Untersuchungen über den postmortalen Liquor-pH in Beziehung zur Leichenzeit, Zacchia, 21, 1, 1958. 66. Evans, W.E.D., The chemistry of death, Charles C. Thomas, Springfield, IL, 1963. 67. Mittmeyer, H.J., Elektrophoretische Untersuchungen über proteolytische Veränderungen an menschlichen Geweben: Ein Beitrag zur Eingrenzung der Todeszeit im späten postmortalen Intervall, Habil.Schrift, Tübingen, 1978. 68. Bonte, W. and Rustemeyer, J., [Quantitative investigation of the amino acid levels in putrefying liver], Z. Rechtsmed., 76, 293, 1975. 69. Mittmeyer, H.J. and Welte, R., [Determination of the time of death after dismemberment. Muscle electrophoretic criteria for estimating the early postmortal period in cadaver parts], Z. Rechtsmed., 88, 23, 1982. 70. Mittmeyer, H.J., [Muscle electrophoretic study in the determination of the time of death], Beitr. Gerichtl. Med., 38, 177, 1980. 71. Lindlar, F., [Postmortem lipid changes and time of death determination], Beitr. Gerichtl. Med., 26, 71, 1969. 72. Döring, G., [Postmortism lipid metabolism], Beitr. Gerichtl. Med., 33, 76, 1975. 73. Bonte, W., Der postmortale Proteinkatabolismus. Experimentelle Untersuchungen zum Problem der forensischen Leichen-zeitbestimmung, Habil. Schrift, Göttingen, 1978. 74. Schmidt, O., Forster, B., and Schulz, G., Untersuchungen über die Anteile der Eigen—und Fremdfermente am postmortalen Eiweisszerfall, Dtsch. Z. ges gerichtl Med., 52, 28, 1961. 75. Bonte, W., et al., [The effect of microorganisms on protein catabolism in putrefaction studies], Beitr. Gerichtl. Med., 34, 173, 1976. 76. Mittmeyer, H.J. and Strebel, K.H., [Experimental examinations on forensic determination of time of death by electrofocusing of soluble muscle protein], Z. Rechtsmed., 85, 235, 1980. 77. Mittmeyer, H.J. and Erlinger, R., [Postmortem proteolysis of human myofibrillar proteins], Beitr. Gerichtl. Med., 37, 291, 1979. 78. Vass, A.A., et al., Decomposition chemistry of human remains: a new methodology for determining the postmortem interval, J. Forensic Sci., 47, 542, 2002. 79. Petroff, O.A., Ogino, T., and Alger, J.R., High-resolution proton magnetic resonance spectroscopy of rabbit brain: regional metabolite levels and postmortem changes, J. Neurochem., 51, 163, 1988.
208
80. Michaelis, T., Helms, G., and Frahm, J., Metabolic alterations in brain autopsies: proton NMR identification of free glycerol, NMR Biomed., 9, 121, 1996. 81. Choe, B.Y., et al., Postmortem metabolic and morphologic alterations of the dog brain thalamus with use of in vivo 1H magnetic resonance spectroscopy and electron microscopy, Invest Radiol., 30, 269, 1995. 82. Kauppinen, R.A., et al., Quantitative analysis of 1H NMR detected proteins in the rat cerebral cortex in vivo and in vitro, NMR Biomed., 6, 242, 1993. 83. Higuchi, T., et al., Mapping of cerebral metabolites in rats by 1H magnetic resonance spectroscopic imaging. Distribution of metabolites in normal brain and postmortem changes, NMR Biomed., 6, 311, 1993. 84. Fineschi, V., et al., 1H-NMR studies of postmortem biochemical changes in rat skeletal muscle, Forensic Sci. Int., 44, 225, 1990. 85. Alanen, A.M., et al., The effects of the method of death and lapsed time on proton relaxation time T1 in autopsied muscle samples, Invest Radiol., 28, 529, 1993. 86. Pearson, R.T., et al., An NMR investigation of rigor in porcine muscle, Biochim. Biophys. Acta, 362, 188, 1974. 87. Chang, D.C., Hazlewood, C.F., and Woessner, D.E., The spin-lattice relaxation times of water associated with early post mortem changes in skeletal muscle, Biochim. Biophys. Acta, 437, 253, 1976. 88. Harada, H., et al., Identification and quantitation by 1H-NMR of metabolites in animal organs and tissues. An application of NMR spectroscopy in forensic science, Forensic Sci. Int., 24, 1, 1984. 89. Harada, H., Shimizu, H., and Maeiwa, M., 1H-NMR of human saliva. An application of NMR spectroscopy in forensic science, Forensic Sci. Int., 34, 189, 1987. 90. Diessner, H., and Lahl, R., [The post mortem determination of hydrogen-ion concentration in brain tissue homogenate and its relation to cause of death, course of death and time of death in selected autopsy material], Zentralbl. Allg. Pathol., 112, 162, 1969. 91. Daldrup, T., [The kinetics of the postmortal bacterial metabolism of the glutamic acid in brain], Z. Rechtsmed., 86, 195, 1981. 92. Daldrup, T., Die Aminosäuren des Leichengehirns, Enke Verlag Stuttgart, Stuttgart, 1984. 93. Daldrup, T., [Practical experiences with the determination of cadaver age by evaluation of bacterial metabolic products], Z. Rechtsmed., 90, 19, 1983. 94. Kreis, R., Quantitative localized 1H MR spectroscopy for clinical use, Prog. NMR Spectroscopy, 31, 155, 1997. 95. Chang, L., et al., In vivo proton magnetic resonance spectroscopy of the normal aging human brain, Life Sci., 58, 2049, 1996. 96. Banaschak, S., et al., Estimation of postmortem metabolic changes in porcine brain tissue using 1H-MR spectroscopy—preliminary results, Int. J. Legal Med., 119, 77, 2005. 97. Bottomley, P.A., Selective volume method for performing localized NMR spectroscopy, 4.480.228, 1984. 98. Provencher, S.W., Estimation of metabolite concentrations from localized in vivo proton NMR spectra, Magn. Reson. Med., 30, 672, 1993. 99. Slotboom, J., Boesch, C., and Kreis, R., Versatile frequency domain fitting using time domain models and prior knowledge, Magn. Reson. Med., 39, 899, 1998.
© 2009 by Taylor & Francis Group, LLC
The Virtopsy Approach
100. Mierisova, S. and Ala-Korpela, M., MR spectroscopy quantitation: a review of frequency domain methods, NMR Biomed., 14, 247, 2001. 101. Vanhamme, L., et al., MR spectroscopy quantitation: a review of time-domain methods, NMR Biomed., 14, 233, 2001. 102. Ith, M., et al., Observation and identification of metabolites emerging during postmortem decomposition of brain tissue by means of in situ 1H-magnetic resonance spectroscopy, Magn. Reson. Med., 48, 915, 2002. 103. Scheurer, E., et al., Statistical evaluation of time-dependent metabolite concentrations: estimation of post-mortem intervals based on in situ 1H-MRS of the brain, NMR Biomed., 18, 163, 2005. 104. Henssge, C. and Madea, B., Estimation of the time since death, Forensic Sci. Int., 165, 182, 2007.
D2.3 VITAL REACTIONS AND VITAL SIGNS Danny Spendlove, Stephan A. Bolliger, and Michael J. Thali
D2.3.1 INTRODUCTION One of the most important questions a forensic pathologist must address is whether a person was already dead or still alive upon infliction of injuries, as this is of great relevance in court as well as for insurance companies. For example, if a driver dies due to a cardiac arrest at the wheel and then crashes his or her car, the manner of death will be natural, or if a car runs over a dead body, then the driver cannot be held responsible for the death of the person [1–4]. This distinction is possible due to the presence or absence of so-called vital signs. These vital signs require some form of reaction of the body toward the inflicting trauma form, toward a circulation, or toward a respiration. Depending on the extent and the type of vital sign, they may also serve to estimate the survival time since infliction of the injury, another question frequently addressed at court. These enormously important signs can be listed as follows [1–5]: Consciousness and reaction of the person toward a trauma Reaction of tissue toward a trauma Blood circulation Respiration and ingestion These are discussed separately in the following sections.
D2.3.2 CONSCIOUSNESS AND REACTION OF A PERSON TOWARD A TRAUMA This is certainly the most obvious vital sign for a medical layperson. The foremost and most frequently encountered sign is the position of the body upon injury infliction. If a pedestrian was walking or standing, then it may be assumed that he or she was alive upon impact of the car. Postmortem imaging can digitalize telltale injuries such as patterned abrasions of the skin by 3D photogrammetry-based
Thanatology
surface scanning and match them to the likewise scanned structures of the vehicle, thus permitting a reconstruction of the impact position of the pedestrian and the vehicle (see also Chapter B1, “External Body Documentation”). Other useful signs indicating an upright position of the pedestrian are wedge fractures (see Chapter D3.2 “Postmortem Imaging of Blunt Trauma”). These lesions arise due to direct impact. In traffic accidents, this implies that due to the distance between the front of the car and the ground, the injured long bone must have been held at least a few centimeters above ground and thus that the person was not lying flat on the ground. On the other hand, defensive injuries to the hands in stabbing incidents imply that the victim either held his or her hands in a defensive manner in front of his or her body or tried to defend himself or herself by grasping the weapon (i.e., the knife blade). However, caution should be exercised in interpreting such defensive findings: Occasionally, natural deaths or suicides are feigned as homicides for insurance purposes. Investigations at the crime scene may show signs that the injured victim crawled away. These findings, although of great forensic importance, belong to the field of crime scene investigation and are therefore not further discussed in detail in this book. In cases of smoke exposure, soot-free lines, crow’s-feetlike lines surround the eyelids as a result of active squinting and are therefore a readily visible vital sign at external inspection of the body [1–4] (Figure D2.3.1).
FIGURE D2.3.1 Elderly man found dead in his burning smokefilled flat. An electrical machine caused a fire in another part of the flat. Note the soot-free lines—crow’s feet lines around the eyes (yellow ring)—indicating that the man squinted. The soot could not settle in the wrinkles. The superficial lesions of the nose and cheek are due to burning of the epidermis.
© 2009 by Taylor & Francis Group, LLC
209
D2.3.3 REACTION OF TISSUE TOWARD A TRAUMA Living tissues strive to repair sustained lesions via an inflammatory response and subsequent repair processes. These reactions—which require a functioning blood circulation— are generally spearheaded by neutrophilic granulocytes and are followed by macrophages and lymphocytes. As these cells are only visible under the microscope, their detection eludes postmortem imaging to date [3,6–9]. In addition to the cellular response to injury the vessel permeability increases, thus giving rise to a perilesional edema not seen in postmortem injuries. A hemorrhage may also occur, visible or nonvisible to the examiner, depending on the site and depth of the bleeding. Postmortem MRI suffices in detecting edema and hemorrhage and thus due to the edema can help in assessing the vitality of the lesion (Figure D2.3.2). However, putrefaction leads to a general extravascular, interstitial fluid accumulation, thus obscuring this accompanying phenomenon of inflammatory repair and rendering the magnetic resonance imaging (MRI) examination regarding the vital edema useless. The appearance of interstitial fluid accumulation and edema may be similar; in the future, differentiation of the fluids by MR sequences should be possible. On the other hand, multislice computed tomography (MSCT) can be of assistance in gaining histological specimens of regions of interest by image-guided biopsy (See Chapter D5, “Biopsy”).
FIGURE D2.3.2 In this total imaging matrix (TIM) MRI image a hyperintense area is shown, which marks a hematoma at the infraspinatic muscle.
210
The Virtopsy Approach
With the development of new MRI and CT scanners, or by labeling antineutrophilic granulocyte antibodies and administering these into the bloodstream with a heart-lung machine (see Chapter D6, “Angiography”), these inflammatory responses can possibly be visualized by postmortem imaging in the near future.
D2.3.4 BLOOD CIRCULATION Here, two different mechanisms regarding the blood flow can be utilized for the assessment of vital reactions: (1) a vascular occlusion due to thrombembolism; and (2) a vascular injury with subsequent bleeding [2,3]. Emboli require a blood flow to reach their destination and are therefore very useful indicators for the vitality of a lesion. Apart from thrombemboli into various organ systems, which may well be lethal, a multitude of other embolus forms exists. Of these, fat, gas, bone marrow, amniotic fluid, and tissue embolisms are the most frequent [2–4,10]. Other, albeit rarer, embolisms are foreign-body embolisms such as projectiles (Figure D2.3.3) and osteocement [11,12]. The fat embolism is indeed the most frequently encountered embolism form in forensic pathology [1–3,13]. It should
not be confused with the intracapillary fat globules seen in burned bodies, which are pressed into the vascular system due to the pressure increase of the heated body water. In even minor blunt trauma, subcutaneous fat lobules are crushed, and their contents are flushed into the venous system and ultimately become lodged in the pulmonary capillaries where they can be detected. To a lesser extent it serves as a reliable vital sign; in severe cases it may even cause death due to right-heart failure. In conventional autopsies, the detection of fat embolism is performed by microscopic examination of cryo-cut, sudan red-stained slides or of unfixed thin lungtissue slices gained by a double-bladed knife stained with sudan red (Figure D2.3.4). These (micro-) emboli cannot be detected yet by postmortem imaging. However, with microscopic examination of image-guided lung-tissue sampling by biopsy, this shortcoming may be solved in a minimally invasive fashion in the near future. The second most frequently seen embolism form is the gas embolism. The gas, usually air, reaches the vascular system via injured veins located above the level of the heart and is then pumped into the system. Such a gas embolism may constitute the cause of death by right-heart failure. It should not be mistaken with gas seen in putrefaction, which is obviously
A
C
B
D
FIGURE D2.3.3 A young woman was executed by several gunshots. One projectile had entered the body in the thorax front, left from the right nipple. The bloodstream transported the projectile to the lumbar aorta, where circulation stopped. The multiplanar reconstruction (MPR) images (A,B) show the projectile in the aorta from different views. A 3D reconstruction with computer-aided coloring of the metal shows the blue projectile (C,D).
© 2009 by Taylor & Francis Group, LLC
Thanatology
FIGURE D2.3.4 Histological lung specimen on a traffic accident victim stained with sudan red, which dyes fat red. Note the innumerous capillaries filled with embolized fat (magn. 20x HE-stain).
a purely postmortem phenomenon. Although in putrefaction the allocation of the gas is seen in all organs, in gas embolism the gas is more or less arranged in the vessels [14–19]. Diagnosing gas embolism with conventional methods is difficult: The heart is punctured while submerged in water in the pericardial sack. If bubbles rise to the water surface, this confirms the presence of gas in the cardiac chamber. In both locations, autopsy is at a great disadvantage: It cannot determine the amount of gas (Figure D2.3.5). Furthermore, the mere presence of gas in a cardiac chamber does not necessarily imply that this gas was embolized. MSCT can overcome these difficulties easily. Sectional images depict the presence of gas immediately (Figure D2.3.6), while 3D reconstructions display the gas distribution in the blood vessels and the cardiac chambers; the precise amount may even be determined (Figure D2.3.7). In contrast to the fat embolism, gas embolism is easily detected by postmortem MSCT [14–19]. After organ crushing (e.g. liver, bone marrow), so-called tissue embolisms can embolize into the pulmonary or other vascular systems. Depending on their size, these embolisms can be seen in MRI. On the other hand, MSCT is better suited for the detection of foreign body, especially metallic emboli (Figure D2.3.8). Injuries invariably cause damage to the vascular system, thus giving rise to a hemorrhage. Although small hemorrhages can arise after cessation of the blood flow by a passive
© 2009 by Taylor & Francis Group, LLC
211
FIGURE D2.3.5 Autoptic examination technique in cases of suspected gas embolism. The pericardial sack is opened and filled with water. The immersed right cardiac chamber is then punctured. If bubbles rise, gas was present in the heart.
leakage of blood from the injured vessels, large hematomas require some form of pressure within the vascular system and are therefore usually vital. Such hematomas are easily seen in postmortem MRI and, to a lesser extent, in postmortem MSCT [28–30] (Figure D2.3.9). Hemorrhages into body cavities, which are easily detected by both cross-sectional modalities, can—depending on their extent—be seen as vital reactions. If
FIGURE D2.3.6 MSCT, axial image of the chest of a homicide victim who was stabbed to death. Note the massive accumulation of gas, in this case air, in the right cardiac chamber (arrow).
212
The Virtopsy Approach
a body cavity (Figure D2.3.10) or by demonstrating that the large blood vessels such as the aorta, which obviously lack blood, have collapsed (see Chapter D3.6, “Fatal Hemorrhage in Postmortem Cross-Sectional Radiology”) [28,29].
D2.3.5 RESPIRATION AND INGESTION Solids, fluids, and gases from within the body as well as from the surrounding environment can enter the respiratory and digestive tract. In the respiratory tract, the ventilation of the lungs leads to these substances entering the bronchioles and, if they are sufficiently small, even the alveoli, where they can enter the bloodstream. Thus, the presence of such substances in the lung periphery or even the blood indicates that the victim was alive when he or she was in direct contact with them [2–4]. Gases such as carbon monoxide and cyanide, both produced in fires, can enter the bloodstream if a ventilation and a circulation is present. The presence of these gases therefore constitutes a vital reaction [31–36]. However, although the suspicion arises at external inspection due to pink livores, detection is based on chemical analyses, which is why they are not further dealt with here. Aspiration of blood is a common finding in cases of skull base trauma and injury to the upper and lower airways and can cause death due to asphyxia [3,37,38]. A deep blood
FIGURE D2.3.7 MSCT, sagittal image (maximum intensity projection) of the head of the victim seen in Figure D2.3.5. The cerebral vessels are filled with gas, thus proving a (lethal) gas embolism.
a large amount blood is lost, this may lead to death. At external inspection, the paucity of livores gives rise to the suspicion of such an event. Both postmortem MSCT and MRI can confirm this diagnosis rapidly by showing the blood accumulation in
A
B
C
D
FIGURE D2.3.8 Suicide by headshot in the temporal bone. The MPR images show the location of the fragments. Notice the small, blue colored, radio-opaque structures on the left side of the cervical spine (projectile fragments).
© 2009 by Taylor & Francis Group, LLC
Thanatology
213
A
B
C
D
FIGURE D2.3.9 A 45-year-old male was run over by a car while on his bike. He was hit on the right side at hip height. The hematoma is visible in both CT (A,B) but is more visible in the MRI image (C). In autopsy we found the hematoma (D).
aspiration into the lung periphery indicates that at least a few breaths took place after the injury infliction and gives rise to the leopard-skin-like texture of the lungs seen at autopsy. Both MRI and MSCT can detect these spots in the lungs of fresh corpses reliably as hyperdense areas in the lungs
FIGURE D2.3.10 MSCT image of a male who was killed in a motorbike accident. He died of exsanguination, represented in the picture by a collapsed aorta and the hematothorax on the right side.
© 2009 by Taylor & Francis Group, LLC
(Figure D2.3.11). Depending on the position at the time of aspiration, different lung regions are predominantly affected: The upper lobes and the superior segments of the lower lobes are affected when the aspiration occurred in a supine position; the lower lobes on both sides are of interest when people aspirated in an upright position. However, in cases of advanced putrefaction the image becomes obscured by the putrefactive fluids, and the diagnosis may become difficult or even impossible to make. Often, gastric contents—usually composed of fluids, small food morsels, or even larger pieces of food—are aspirated [39,40]. Differentiation among a vital aspiration (i.e., antemortem), an agonal (i.e., perimortem) aspiration, and a postmortem influx due to turning of the corpse during the medicolegal examination must be made. Vital gastric content aspirations generally occur in persons in which the glottis reflex is impaired, such as in craniocerebral injuries with coma or in instances of intoxication. Here, depending on the size, the contents are aspirated into the lung periphery. Often the mucosal lining of the airways displays an inflammatory response to the acidic gastric fluid. The agonal aspiration of gastric contents is frequently seen. By contrast to the truly vital aspirations, the contents do not reach the lung periphery. Obviously, an inflammatory response is missing. MSCT and MRI can depict the morsels within the airways fairly easily (Figure D2.3.12). As opposed to blood aspiration, they are usually not round hyperdense areas but are rather polymorph
214
The Virtopsy Approach
A
B
C
D
FIGURE D2.3.11 (A): MSCT view on blood aspiration in the lungs. (B): MRI image, blood aspiration. (C): Bloodaspiration at autopsy. (D): The aspiration in MSCT, shown in (C).
regions. Another frequently aspirated fluid is water in cases involving drowning. As this is discussed in detail in Chapter D3.8, “Drowning: Postmortem Imaging Findings,” this is not further dealt with here. In cases of fire, soot may also be aspirated into the lungs [2–4,41]. The detection of soot in the respiratory tract therefore implies that the person breathed smoke and was therefore
alive when the fire broke out (Figure D2.3.13). However, the detection of soot in the airways is not yet possible with MSCT or MRI. Nevertheless, a minimally invasive approach in determining whether a person was alive during the outbreak of a fire can be made by detecting the carbon monoxide and cyanide concentrations in the blood, the crow’s feet lines around the eyelids, and a laryngoscopy. If a victim was
A
B
C
D
FIGURE D2.3.12 (A,C): Gastric content aspiration, seen on the two MSCT images. (B,D): MRI images showing gastric content aspiration in the lungs.
© 2009 by Taylor & Francis Group, LLC
Thanatology
FIGURE D2.3.13A Autopsy image of the opened lower airways. Note the soot in the bronchi (arrows).
buried alive, aspirated earth or sand may be detected in the airways [42–46]. As in cases of soot aspiration, postmortem imaging fails to detect such material unless the soot fills out the buccal cavity. Therefore, an inspection of the mouth and perhaps a laryngoscopic examination may solve such cases in a minimally invasive manner. Besides the aforementioned, ingestion of certain substances into the digestive tract also has implications regarding the vitality of a lesion. Blood not only can be aspirated but also can be swallowed. Although a passive, postmortem flowing of blood into the stomach is occasionally seen, the passage into the duodenum or even more distal parts of the
FIGURE D2.3.13B Axial MSCT view of a fire victim. The thick soot layer is visible in the bronchus, marked by the red arrow.
© 2009 by Taylor & Francis Group, LLC
215
FIGURE D2.3.14 MRI image of a homicide victim who was shot through the base of the skull and died of asphyxia due to blood aspiration. The blood is seen in the stomach, which can get there only if blood was actively swallowed.
digestive tract requires peristaltic movements and indicates that the person was alive at the time of injury infliction. This presence of blood in the gastrointestinal tract (see also Chapter D.3.1.3, “Other Organ-System Pathologies”) can be visualized with both MSCT and MRI (Figure D2.3.14). The distance the blood passed through the digestive tract can help give a cautious estimation of the survival time. Obviously, other objects can also be swallowed. Besides soot, solid objects such as teeth, bone fragments, and glass have been found in the esophagus and stomach (Figure D2.3.15). The passage of such solid objects through the esophagus
FIGURE D2.3.15 Axial MSCT of the chest of a traffic victim. A small, radio-opaque structure (arrow) is seen lying within the esophagus. The insert depicts the corresponding autopsy photograph: Here a small piece of glass is shown.
216
The Virtopsy Approach
implies an active transport (i.e., swallowing) and therefore that the victim was alive.
D2.3.6 CONCLUSION Postmortem imaging with MSCT and MRI can detect certain forms of vital signs: Gas embolism with MSCT Blood and stomach content aspiration (if the corpse is not decomposed) in MSCT and MRI Ingested blood in MSCT and MRI Swallowed foreign particles with MSCT Large hemorrhages and hematomas in MRI and to a lesser extent in MSCT Perilesional edemas in MRI However, they do not suffice yet to detect the following: Fat embolism Soot aspiration or ingestion Inhaled gases Soot-free wrinkles (crow’s feet lines) on the face Apart from crow’s feet lines, which are not detected at autopsy but rather at external inspection, the only vital reaction seen at autopsy but not with postmortem imaging is soot aspiration. This shortcoming can be corrected with the use of a laryngoscope, with which soot in the airways can be seen easily. Fat embolism is not seen at autopsy but in the subsequent histological specimens of the lungs. This can be performed in a minimally invasive fashion, too, as the performing of MSCT-image-guided biopsy will also deliver the necessary specimens for microscopic examination. The same goes for the chemical detection of inhaled gases: MSCT-guided sampling of blood from the cardiac chambers will allow for the necessary chemical examination. However, with postmortem imaging an extremely important vital sign not clearly seen at autopsy can be proven: gas embolism. The current autopsy technique for the examination of a possible gas embolism is the opening of the submersed heart. With this method, only gas within the cardiac chambers—but not necessarily an embolization—can be shown. Therefore, we conclude that postmortem imaging, in addition to laryngoscopy and biopsy, is a viable tool for the minimally invasive detection of vital signs.
D2.3.7 REFERENCES 1. DiMaio, V.J. and D. DiMaio. 2001. Electrocution, in Forensic Pathology, 2d. ed. CRC Press, London. 2. Saukko, P. and B. Knight. 2004. Knights Forensic Pathology, 3d ed. Arnold Publishers, Oxford University Press, New York.
© 2009 by Taylor & Francis Group, LLC
3. Brinkmann, B. and B. Madea. 2004. Handbuch gerichtliche Medizin [Handbook of forensic medicine, authors’ translation]. Springer Verlag, Berlin. 4. Reimann,W. and O. Prokop. 1985. Vademecum Gerichtsmedizin, 4th ed. [Vademecum forensic medicine, authors’ translation]. VEB Verlag Volk und Gesundheit, Berlin. 5. Thali, M.J., K. Yen, W. Schweitzer, et al. 2003. Virtopsy, a new imaging horizon in forensic pathology: virtual autopsy by postmortem multislice computed tomography (MSCT) and magnetic resonance imaging (MRI)—a feasibility study. J Forensic Sci 48(2):386–403. 6. Hernandez-Cueto, C., E. Girela, and D.J. Sweet. 2000. Advances on the diagnosis of wound vitality: a review. Am J Forensic Med Pathol 21(1):21–31. 7. Rajs, J. and I. Thiblin. 2000. Histologic appearance of fractured thyroid cartilage and surrounding tissues. Forensic Sci Int 128(1-2):24–28. 8. Oehmichen, M. 2004. Vitality and time course of wounds. Forensic Sci Int 144(2–3):221–31. 9. Takabe, F. and N. Fujitani. 1984. Novel methods for determining the vital reaction in injured skin. Ann Acad Med Singapore 13(1):69–76. 10. Rainio, J. and A. Penttila. 2003. Amniotic fluid embolism as cause of death in a car accident—a case report. Forensic Sci Int 137(2–3):231–34. 11. Pollak, S., D. Ropohl, and M. Bohnert. 1999. Pellet embolization to the right atrium following double shotgun injury. Forensic Sci Int 99(1):61–69. 12. Dada, M.A., I.A. Loftus, and G.S. Rutherfoord. 1993. Shotgun pellet embolism to the brain. Am J Forensic Med Pathol 14(1):58–60. 13. Reh, H. 1980. [Fat and bone marrow embolism as a vital reaction]. Beitr Gerichtl Med 38:147–53. 14. Bajanowski, T., A. West, and B. Brinkmann. 1998. Proof of fatal air embolism. Int J Legal Med 111(4):208–11. 15. Bunai Y., A. Nagai, I. Nakamura, S. Kanno, S. Yamada, and I. Ohya. 1999. An unusual case of fatal gas embolism. Am J Forensic Med Pathol 20(3):256–60. 16. Thali, M.J., K. Yen, P. Vock, et al. 2003. Image-guided virtual autopsy findings of gunshot victims performed with multi-slice tomography and magnetic resonance imaging and subsequent correlation between radiology and autopsy findings. Forensic Sci Int 138(1-3):8–16. 17. Jackowski, C., M. Thali, M. Sonnenschein, et al. 2004. Visualization and quantification of air embolism structure by processing postmortem MSCT data. J Forensic Sci 49(6):1339–42. 18. Ozdoba, C., J. Weis, T. Plattner, R. Dirnhofer, and K. Yen. 2005. Fatal scuba diving incident with massive gas embolism in cerebral and spinal arteries. Neuroradiology 49(6):411–16. 19. Aghayev, E., K. Yen, M. Sonnenschein, et al. 2005. Pneumomediastinum and soft tissue emphysema of the neck in postmortem CT and MRI; a new vital sign in hanging? Forensic Sci Int 153(2–3):181–88. 20. Perrot, L.J. and R.C. Froede. 1985. Bone marrow emboli versus fat emboli as the cause of unexpected death. J Forensic Sci 30(2):338–44. 21. Huber, A., A. Dorn, A. Witzmann, and J. Cervos-Navarro. 1993. Microthrombi formation after severe head trauma. Int J Legal Med 106(3):152–55. 22. Miyaishi, S., F. Moriya, Y. Yamamoto, and H. Ishizu. 1994. Massive pulmonary embolization with cerebral tissue due to gunshot wound to the head. Brain Inj 8(6):559–64.
Thanatology
23. Lau, G. 1995. Pulmonary cartilage embolism: fact or artefact? Am J Forensic Med Pathol 16(1):51–53. 24. Michalodimitrakis, M. and A. Tsatsakis. 1998. Massive pulmonary embolism by liver tissue. Med Sci Law 38(1):85–87. 25. Rainio, J. and A. Penttila. 2003. Amniotic fluid embolism as cause of death in a car accident—a case report. Forensic Sci Int 137(2–3):231–34. 26. Morentin, B. and B. Biritxinaga. 2006. Massive pulmonary embolization by cerebral tissue after head trauma in an adult male. Am J Forensic Med Pathol 27(3):268–70. 27. Thali, M.J., K. Yen, T. Plattner, et al. 2002. Charred body: virtual autopsy with multi-slice computed tomography and magnetic resonance imaging. J Forensic Sci 47(6):1326–31. 28. Aghayev, E., M. Sonnenschein, C. Jackowski, et al. 2006. Postmortem radiology of fatal hemorrhage: measurements of cross-sectional areas of major blood vessels and volumes of aorta and spleen on MDCT and volumes of heart chambers on MRI. AJR Am J Roentgenol 187(1):209–15. 29. Aghayev E., C. Jackowski, M. Sonnenschein, M. Thali, K. Yen, and R. Dirnhofer. 2006. Virtopsy hemorrhage of the posterior cricoarytenoid muscle by blunt force to the neck in postmortem multislice computed tomography and magnetic resonance imaging. Am J Forensic Med Pathol 27(1):25–29. 30. Jackowski, C., M. Thali, E. Aghayev, et al. 2006. Postmortem imaging of blood and its characteristics using MSCT and MRI. Int J Legal Med 120(4):233–40. 31. Sigrist, T. and R. Dirnhofer. 1979. Zur Entstehung der kombinierten inhalatorischen Blausäure-KohlenmonoxidVergiftung. Arch Kriminol 163:145–59. 32. Kojima, T., M. Yashiki, F. Chikasue, and T. Miyazaki. 1990. Analysis of inflammable substances to determine whether death has occurred before or after burning. Z Rechtsmed 103(8):613–19. 33. Rogde, S. and J.H. Olving. 1996. Characteristics of fire victims in different sorts of fires. Forensic Sci Int 77(1–2):93–99.
© 2009 by Taylor & Francis Group, LLC
217
34. Wirthwein, D.P. and J.E. Pless. 1996. Carboxyhemoglobin levels in a series of automobile fires. Death due to crash or fire? Am J Forensic Med Pathol 17(2):117–23. 35. Jaffe, F.A. 1997. Review: Pathogenicity of carbon monoxide. Am J Forensic Med Pathol 18(4):406–10. 36. Abu-al Ragheb, S.Y. and A.H. Battah. 1999. Carbon monoxide fatalities in medicolegal autopsies. Med Sci Law 39(3):243–46. 37. Yen, K., T. Plattner, and R. Dirnhofer. 2005. Retrograde blood aspiration: a vital reaction. Forensic Sci Int 154(1):13–18. 38. Tsokos, M. and R.W. Byard. 2007. Massive, fatal aspiration of blood: not necessarily a result of trauma. Am J Forensic Med Pathol 28(1):53–54. 39. Knight, B. 1975. The significance of the pulmonary discovery of gastric contents in the airway passages. Forensic Sci 6:229–34. 40. Nimmo, W.S. 1985. Aspiration of gastric contents. Br J Hosp Med 34(3):176–79. 41. Suzuki T., H. Takahashi, and K. Umetsu. 1995. Unusual aspirations in fire death. Forensic Sci Int 72(1):71–76. 42. Choy, I.O. and O. Idowu. 1996. Sand aspiration: a case report. J Pediatr Surg 31(10):1448–50. 43. Dunagan, D.P., J.E. Cox, and M.C. Chang. 1997. Haponik EF. Review: Sand aspiration with near-drowning. Radiographic and bronchoscopic findings. Am J Respir Crit Care Med 156(1):292–95. 44. Glinjongol, C., S. Kiatchaipipat, and S. Thepcharoenniran. 2004. Severe sand aspiration: a case report with complete recovery. J Med Assoc Thai 87(7):825–28. 45. Hanson, K.A., J.D. Gilbert, R.A. James, and R.W. Byard. 2002. Upper airway occlusion by soil—an unusual cause of death in vehicle accidents. J Clin Forensic Med 9(2):96–99. 46. Maxeiner, H. and V. Schneider. 1985. [Suffocation death by occlusion of the airways with sand]. Z Rechtsmed 94(3):173–89.
D3
Incident-Specific Cases
CONTENTS D3.1 Natural Death: Introduction ........................................................................................................................................... 222 D3.1.1 Cerebral Pathology in Natural Death ............................................................................................................... 222 D3.1.1.1 Introduction ..................................................................................................................................... 222 D3.1.1.2 Brain Mass Alteration ..................................................................................................................... 223 D3.1.1.3 Epileptic Disorders .......................................................................................................................... 224 D3.1.1.4 Vessel Occlusion ............................................................................................................................. 225 D3.1.1.5 Hemorrhagic Lesions ...................................................................................................................... 226 D3.1.1.6 Vascular Malformations .................................................................................................................. 227 D3.1.1.7 Inflammatory Processes ................................................................................................................. 228 D3.1.1.8 Tumorous Lesions ........................................................................................................................... 229 D3.1.1.9 References ....................................................................................................................................... 229 D3.1.2 Cardiac Pathology ............................................................................................................................................ 230 D3.1.2.1 Background ..................................................................................................................................... 230 D3.1.2.2 How to Get Cardiac Cross-Sectional Images ................................................................................. 230 D3.1.2.3 Postmortem Alterations .................................................................................................................. 231 D3.1.2.4 Pathological Findings ..................................................................................................................... 234 D3.1.2.5 Conclusion ...................................................................................................................................... 245 D3.1.2.6 Acknowledgments........................................................................................................................... 246 D3.1.2.7 References ....................................................................................................................................... 246 D3.1.3 Other Organ-System Pathologies ..................................................................................................................... 249 D3.1.3.1 Vascular System .............................................................................................................................. 249 D3.1.3.2 Respiratory System ......................................................................................................................... 250 D3.1.3.3 Gastrointestinal Tract...................................................................................................................... 253 D3.1.3.4 References ....................................................................................................................................... 253 D3.2 Postmortem Imaging of Blunt Trauma .......................................................................................................................... 254 D3.2.1 Introduction ..................................................................................................................................................... 254 D3.2.2 Injury Type, Mechanism, and Cause of Death ............................................................................................... 254 D3.2.2.1 Blunt Trauma of the Superficial Skin Layers ................................................................................. 254 D3.2.2.2 Blunt Trauma of the Subcutaneous Fatty Tissue and Musculature ............................................... 254 D3.2.2.3 Blunt Trauma to the Head .............................................................................................................. 255 D3.2.2.4 Blunt Trauma to the Neck .............................................................................................................. 261 D3.2.2.5 Thorax ............................................................................................................................................ 264 D3.2.2.6 Abdomen and Pelvic Girdle ........................................................................................................... 267 D3.2.2.7 Extremities ..................................................................................................................................... 269 D3.2.3 Conclusion ....................................................................................................................................................... 271 D3.2.4 References........................................................................................................................................................ 271 D3.3 Forensic Neuroimaging .................................................................................................................................................. 272 D3.3.1 Introduction...................................................................................................................................................... 272 D3.3.2 Technical Aspects ............................................................................................................................................ 273 D3.3.2.1 Postmortem In Situ Cranial CT ...................................................................................................... 273 D3.3.2.2 Postmortem In Situ Cranial MRI .................................................................................................. 273 D3.3.2.3 Autopsy ........................................................................................................................................... 274 D3.3.2.4 Image Reading and Forensic Expertise .......................................................................................... 274 D3.3.2.5 Artifact and Limitations ................................................................................................................. 274 D3.3.3 Extracranial Trauma: Scalp and Muscle Tissues, Skull, Upper Cervical Vertebrae....................................... 276 D3.3.3.1 Scalp and Soft Tissues, Temporal Muscles ..................................................................................... 276 D3.3.3.2 Skull and Upper Cervical Vertebrae .............................................................................................. 276 D3.3.4 Traumatic Intracranial Injury .......................................................................................................................... 277 219 © 2009 by Taylor & Francis Group, LLC
220
The Virtopsy Approach
D3.3.4.1 Commotio and Contusio Cerebri ................................................................................................... 277 D3.3.4.2 Diffuse Axonal Injury .................................................................................................................... 277 D3.3.4.3 Extra-Axial Hemorrhage................................................................................................................ 279 D3.3.4.4 Intra-Axial Hemorrhage ................................................................................................................. 284 D3.3.5 Traumatic Injuries: Blunt Force Trauma ........................................................................................................ 288 D3.3.6 Traumatic Injuries: Penetrating Injuries .......................................................................................................... 289 D3.3.6.1 Gunshot Injuries.............................................................................................................................. 289 D3.3.6.2 Penetrating Injuries: Others ............................................................................................................ 290 D3.3.6.3 Brain Laceration ............................................................................................................................. 290 D3.3.6.4 Decerebration ................................................................................................................................. 291 D3.3.7 Embolism: Gas, Fat, Foreign Bodies ............................................................................................................... 291 D3.3.7.1 Gas Embolism ................................................................................................................................. 291 D3.3.7.2 Cerebral Fat Embolism ................................................................................................................... 291 D3.3.7.3 Foreign Bodies ................................................................................................................................ 291 D3.3.8 Nontraumatic Injury ........................................................................................................................................ 291 D3.3.8.1 Cerebrovascular Diseases ............................................................................................................... 291 D3.3.8.2 Neoplasm ........................................................................................................................................ 293 D3.3.8.3 Inflammation .................................................................................................................................. 294 D3.3.8.4 Hereditary Malformation .............................................................................................................. 297 D3.3.8.5 Degenerative Alterations, “White Matter Lesions”........................................................................ 297 D3.3.9 Toxicological Aspects ...................................................................................................................................... 299 D3.3.9.1 Changes in Chronic Alcohol Abuse................................................................................................ 299 D3.3.9.2 Methanol ......................................................................................................................................... 299 D3.3.9.3 Drug Abuse ..................................................................................................................................... 299 D3.3.9.4 Carbon Monoxide Intoxication ....................................................................................................... 299 D3.3.10 Hypoxic Encephalopathy ............................................................................................................................... 299 D3.3.11 Edema and Increased Brain Pressure ............................................................................................................ 299 D3.3.11.1 Edema ............................................................................................................................................ 299 D3.3.11.2 Brain Herniation............................................................................................................................ 300 D3.3.12 Brain Death .................................................................................................................................................... 300 D3.3.12.1 “Respirator Brain” ........................................................................................................................ 300 D3.3.13 Decomposition ............................................................................................................................................... 301 D3.3.14 Clinical Forensic Neuroimaging .................................................................................................................... 301 D3.3.15 Future Aspects ............................................................................................................................................... 301 D3.3.16 Acknowledgments .......................................................................................................................................... 302 D3.3.17 References ...................................................................................................................................................... 302 D3.4 Sharp Trauma ................................................................................................................................................................ 304 D3.4.1 Introduction...................................................................................................................................................... 304 D3.4.2 Stab Wounds .................................................................................................................................................... 304 D3.4.3 Cuts ...................................................................................................................................................................311 D3.4.4 Chop and Hacking Wounds ..............................................................................................................................314 D3.4.5 Conclusion ........................................................................................................................................................317 D3.4.6 References .........................................................................................................................................................317 D3.5 Gunshot ...........................................................................................................................................................................318 D3.5.1 Introduction.......................................................................................................................................................318 D3.5.2 General Ballistics .............................................................................................................................................318 D3.5.3 General Classification of Gunshot Wounds ......................................................................................................319 D3.5.4 Forensic Aspects of Gunshot Injuries ...............................................................................................................319 D3.5.4.1 Weapon Handling ............................................................................................................................319 D3.5.4.2 Firing Distance ................................................................................................................................319 D3.5.4.3 Penetrating or Grazing Injury......................................................................................................... 320 D3.5.4.4 Entrance and Exit Wounds ............................................................................................................. 321 D3.5.4.5 Bullet Course through the Body ..................................................................................................... 323 D3.5.4.6 Gunshot Priority ............................................................................................................................. 326 D3.5.4.7 Lodging of the Bullet or Other Foreign Bodies .............................................................................. 327 D3.5.4.8 Cause of Death and Vitality of the Injuries .................................................................................... 328 D3.5.4.9 Bullet Type and Size, Identification of the Individual Weapon ...................................................... 328
© 2009 by Taylor & Francis Group, LLC
Incident-Specific Cases
221
D3.5.5 Head Injuries .................................................................................................................................................. 328 D3.5.6 Conclusion .......................................................................................................................................................331 D3.5.7 References .......................................................................................................................................................331 D3.6 Fatal Hemorrhage in Postmortem Cross-Sectional Radiology .....................................................................................331 D3.6.1 Clinical Radiological Experience .................................................................................................................. 332 D3.6.2 Postmortem Imaging Findings in Cases Who Died Due to Fatal Hemorrhage............................................. 333 D3.6.3 References ...................................................................................................................................................... 334 D3.7 Strangulation ................................................................................................................................................................ 334 D3.7.1 Strangulation by Hanging ............................................................................................................................... 335 D3.7.2 Manual Strangulation ..................................................................................................................................... 335 D3.7.3 References....................................................................................................................................................... 337 D3.8 Drowning: Postmortem Imaging Findings................................................................................................................... 338 D3.8.1 Introduction ................................................................................................................................................... 338 D3.8.2 Typical Signs of Drowning in Postmortem Imaging ..................................................................................... 338 D3.8.2.1 Aspiration ...................................................................................................................................... 338 D3.8.2.2 Emphysema Aquosum ................................................................................................................... 339 D3.8.2.3 Heart Failure.................................................................................................................................. 339 D3.8.2.4 Pulmonary Edema ......................................................................................................................... 339 D3.8.2.5 Hemodilution ................................................................................................................................. 339 D3.8.2.6 Distension of Stomach and Duodenum ......................................................................................... 339 D3.8.2.7 Water in the Paranasal Sinuses ...................................................................................................... 339 D3.8.3 Conclusion ...................................................................................................................................................... 340 D3.8.4 References ...................................................................................................................................................... 342 D3.9 Thermal Damage .......................................................................................................................................................... 343 D3.9.1 Heat and Burns ............................................................................................................................................... 343 D3.9.1.1 Introduction .................................................................................................................................. 343 D3.9.1.2 Stages of Burn Wounds ................................................................................................................ 343 D3.9.1.3 Area of Involvement ..................................................................................................................... 344 D3.9.1.4 Postmortem Heat Injury ............................................................................................................... 345 D3.9.1.5 Vital Reactions ............................................................................................................................. 346 D3.9.1.6 Forensic Examination .................................................................................................................. 346 D3.9.1.7 Identification ................................................................................................................................. 348 D3.9.1.8 Vital Signs during Fire Exposure in CT ...................................................................................... 350 D3.9.1.9 Alcohol, Drugs, or Other Burn-Specific Analyses (Carbon Monoxide, Hydrogen Cyanide from Burning Polymers)............................................................................................................... 352 D3.9.1.10 Cause of Death and Injuries ......................................................................................................... 352 D3.9.1.11 References..................................................................................................................................... 354 D3.9.2 Hypothermia................................................................................................................................................... 354 D3.9.2.1 References ...................................................................................................................................... 356 D3.10 Electricity ..................................................................................................................................................................... 357 D3.10.1 Introduction .................................................................................................................................................. 357 D3.10.1.1 Effect of Frequency ...................................................................................................................... 357 D3.10.1.2 Thermal Effects............................................................................................................................ 358 D3.10.1.3 Sensation ...................................................................................................................................... 358 D3.10.1.4 Motor Function ............................................................................................................................. 358 D3.10.1.5 Cardiac Function ......................................................................................................................... 358 D3.10.1.6 Nerve Function ............................................................................................................................. 358 D3.10.1.7 Burns ............................................................................................................................................ 358 D3.10.2 Risk Factors .................................................................................................................................................. 358 D3.10.3 Forensic Findings ........................................................................................................................................ 358 D3.10.4 Radiological Findings ................................................................................................................................... 360 D3.10.5 Micro-MR Findings ..................................................................................................................................... 361 D3.10.6 Conclusion .................................................................................................................................................... 362 D3.10.7 References ..................................................................................................................................................... 362 D3.11 Clinical Forensic Imaging ............................................................................................................................................ 363 D3.11.1 Introduction................................................................................................................................................... 363 D3.11.2 Standard Forensic Examination of Living Persons ...................................................................................... 364
© 2009 by Taylor & Francis Group, LLC
222
The Virtopsy Approach
D3.11.3 Radiology in Clinical Forensic Medicine .................................................................................................... 364 D3.11.4 Future of Clinical Forensic Radiology: Potential Applications .................................................................. 365 D3.11.4.1 Strangulation ........................................................................................................................... 367 D3.11.4.2 Accidental Trauma................................................................................................................... 368 D3.11.4.3 Child Abuse ............................................................................................................................. 369 D3.11.4.4 Abuse of the Elderly ................................................................................................................ 372 D3.11.4.5 Torture ..................................................................................................................................... 373 D3.11.4.6 Self-Inflicted Injury ................................................................................................................. 373 D3.11.4.7 Body Packing .......................................................................................................................... 373 D3.11.4.8 Sexual Assault ......................................................................................................................... 373 D3.11.4.9 Personal Fitness: Fitness to Drive, Fitness to Act, Fitness to Be Imprisoned ......................... 373 D3.11.4.10 Forensic Psychiatry .................................................................................................................. 373 D3.11.4.11 Civil Law Issues: Claims for Damages .................................................................................... 374 D3.11.4.12 Intoxication .............................................................................................................................. 374 D3.11.4.13 Medical Malpractice ................................................................................................................ 374 D3.11.4.14 Age Estimation ......................................................................................................................... 374 D3.11.5 The Use of Clinical Images for Forensic Purposes...................................................................................... 374 D3.11.6 The Role of Clinical Forensic Imaging in the Juridical Context ................................................................. 375 D3.11.7 Ethical and Juridical Considerations ............................................................................................................ 375 D3.11.8 Conclusion .................................................................................................................................................... 376 D3.11.9 References .................................................................................................................................................... 376 D3.12 Medical Malpractice .................................................................................................................................................... 378 D3.12.1 Introduction .................................................................................................................................................. 378 D3.12.2 The Medical Malpractice Claim ................................................................................................................. 379 D3.12.2.1 Expert Testimony in Malpractice Cases ................................................................................. 379 D3.12.2.2 Forensic Assessment in Medical Malpractice Cases ............................................................... 381 D3.12.2.3 Postmortem CT and MRI in Malpractice Cases: Preliminary Experience ............................. 381 D3.12.2.4 The Use of Radiological Data for Clinical Forensic Case Assessment ................................... 382 D3.12.2.5 Limitations............................................................................................................................... 383 D3.12.3 Conclusion.................................................................................................................................................... 385 D3.12.4 Acknowledgments ........................................................................................................................................ 385 D3.12.5 References .................................................................................................................................................... 385
D3.1 NATURAL DEATH: INTRODUCTION Lars Oesterhelweg and Michael J. Thali A large category of deaths handled in medicolegal investigations encompasses those of natural origin. Within this category are the unexpected deaths of individuals functioning in the community who suddenly collapse and die and individuals who are found dead, such as in their apartment with an unlocked entrance. In some of these cases, superficial injuries such as abrasion or bruises could be found due to so-called agonal injuries, and a foreign hand involved in the cause of death could often not be excluded by scene investigation and external examination. These injuries are often bruises or lacerations on the face or to the galea due to falls or uncoordinated movements in the moment of death that might mimic blunt force from foreign hands. On the other hand, many cases with seemingly natural causes of death (e.g., pneumonia) are directly related to nonnatural triggers like accidents or assaults. The following chapters give an overview of the forensic imaging possibilities in certain natural causes of sudden
© 2009 by Taylor & Francis Group, LLC
death due to cerebral, cardiac, vascular, respiratory, and digestive diseases.
D3.1.1 CEREBRAL PATHOLOGY IN NATURAL DEATH Danny Spendlove, Stephan A. Bolliger, Steffen G. Ross, Andreas Christe, and Michael J. Thali D3.1.1.1 Introduction In forensic pathology, pathologic cerebral findings are routinely encountered. They may be the result of a direct or indirect craniocerebral trauma with minor or extensive injuries to the brain and its coverings, may be the consequence of a systemic reaction such as generalized hypoxia or metabolic disorders, or rarely may arise due to a direct cerebral pathology such as infections, vascular disorders, and tumors. These findings may be concomitant affections to the cause of death or even may present the cause of death. Apart from traumatic lesions, dealt with in detail in Chapter D3.3, “Forensic Neuroimaging,” the following most common,
Incident-Specific Cases
223
nontraumatic intracranial findings and causes of death encountered in forensic pathology [1,2] are discussed in this chapter: r r r r r r r
Brain mass alteration Epileptogenic disorders Blood vessel occlusion Hemorrhagic lesions Vascular malformations Infection Tumorous lesions
D3.1.1.2 Brain Mass Alteration Both cerebral edema and atrophy lead to an alteration of the physiologic brain mass. Although both conditions may be concomitant findings in cross-sectional imaging or autopsy or are case relevant with regard to cerebral function as in cases of progressive dementia. Cerebral edema is a condition that results from a multitude of affections such as hypoxia, congestion, intoxications, or metabolic disorders. Generally, the brain responds to such conditions or lesions in a rather uniform fashion: It swells. This swelling is due to an electrolyte shift into the intracellular space and is easily detected upon autopsy. If the brain swells above a certain degree, the pressure exerted upon the afferent and deferent blood vessels may exceed the physiologic blood pressure, thus giving rise to a collapse in blood circulation of the brain. The result is death due to cerebral hypoxia. This condition is frequently seen in everyday forensic practices; however, great caution should be exercised in interpreting such a finding. Inexperienced pathologists often rely on a
A
higher-than-normal brain weight to diagnose such an edema. This assumption is incorrect because the physiological brain weight of different persons may differ to such an extent that no reliable interindividual assessment is possible. Here, the overall aspect of the brain is far more reliable; if the gyri are flattened and the sulci are thin, then a certain cerebral edema may be postulated. In cases of relevant edema, the uncus and the cerebellar structures may herniate. This increased brain volume resulting in cerebral and cerebellar herination and possible circulatory arrest is easily detected at postmortem imaging in MSCT within the restricted space of the skull (Figure D3.1.1.1, Figure D3.1.1.2, and Figure D3.1.1.3) [3]. The opposite—namely, a cerebral shrinking, or atrophy— is rarely of interest in the forensic evaluation of the cause of death. Nevertheless, it may play an important role in the reconstruction of the case or rather its course of events. Today, with an ever aging population, cerebral atrophy—which is on one hand a physiologic phenomenon of aging and on the other hand is also a sign of degenerative pathologic disorders that often result in dementia—is frequently encountered. The aged brain shows typical alterations that are in most cases not associated with functional impairment. In the postmortem examination of forensic cases, these changes are regularly seen and must be distinguished from case-relevant findings. As in the normal aging brain, several forms of dementia are also accompanied by atrophy of the brain structures that is observed at imaging. Without having information about the clinical status of the person before death, the differential diagnosis between normal aging and pathologic processes will be almost impossible. Alzheimer disease atrophy usually includes both hippocampi and the temporal lobes, whereas in
B 5.5 mm (2D) 5.5 mm (2D)
C
FIGURE D3.1.1.1 Position of the left cerebellar tonsil on sagittal plane of reformatted CT image. Measurement of the position composed 5.5 mm below the foramen magnum line: (A) Whole brain. (B) Enlarged foramen magnum. (C) Predominant swelling of the left cerebellar tonsil in autopsy (arrow).
© 2009 by Taylor & Francis Group, LLC
224
The Virtopsy Approach
A
B
FIGURE D3.1.1.2 (A) Position of left cerebellar tonsil on sagittal plane of T2-weighted MRI. Measurement of the position composed 5.1 mm below the foramen magnum line: (B) Extensively symmetric axial plane of T2-weighted MRI just under the foramen magnum. Predominant herniation of the left cerebellar tonsil (arrows).
Pick’s disease the frontal lobes and frontotemporal regions are the predominantly affected. Although postmortem MSCT and MRI can show signs of atrophy, great caution should be exercised in overinterpreting such findings as signs of dementia, since persons with cerebral atrophy may show no cognitive impairment, and vice versa (Figure D3.1.1.4).
D3.1.1.3 Epileptic Disorders Despite a great deal of research on the cause of death in epileptic persons, many cases remain unsatisfactorily solved [4–6]. Indeed, about 3% to 4% of all natural deaths concern epilepsy-associated deaths. According to Dimaio [1], the incidence of sudden death among epileptics is about 2% to 17%. Possible mechanisms of death are apnea during an epileptic seizure, a neural lung edema, and autonomously induced
A
cardiac arrhythmias [7,8], but very few persons die in status epilepticus [1]. Obviously, epileptics may suffer seizures during certain potentially dangerous activities such as driving and die due to other traumatic causes. In nontraumatic-epilepsy-associated death, the only findings that can imply such an incident are tongue bites, loss of urine, and a scene in acute disarray such as overthrown furniture and disordered blankets. However, these findings are also seen in other settings, such as in violent death, and may lack completely in epilepsy-associated deaths [9] and are therefore not pathognomonic for epileptic seizures. Sometimes, possible causes for seizures such as cerebral sclerosis (scars), dysplastic brain tissue, arteriovenous malformations, or adhesions between cortex and dura and tumors can be found. A subtherapeutic or even lacking blood concentration of anticonvulsive medication in patients with a known history of seizures may give rise to the hypothesis of an epilepsy-associated death. However, these findings do not
B
C 2.7 mm (2D)
2.7 mm (2D)
2.7 mm (2D)
FIGURE D3.1.1.3 Position of the tonsils on sagittal plane of reformatted CT image. Measurement of the position composed on same level as MRI examination, 2.7 mm below the foramen magnum line: (A) Whole brain. (B) Enlarged foramen magnum. (C) Position of the tonsils on sagittal plane of T2-weighted MRI. Measurement of the position composed 2.7 mm below the foramen magnum line.
© 2009 by Taylor & Francis Group, LLC
Incident-Specific Cases
FIGURE D3.1.1.4 Brain edema after intravenous drug abuse. Typical finding.
necessarily prove an epileptic death. Indeed, such a diagnosis is based on the exclusion of other pathological findings and on the patient history. Postmortem imaging can be of assistance in the assessment of such cases in two ways: (1) Traumatic causes of death can be ruled out with MSCT (see Chapter D3.3, “Forensic Nevroimaging); and (2) epiletigenic foci, such as mesio-temporal sclerosis (Figure D3.1.1.5) dysplasia, and arteriovenous malformations (see next section) can be made visible by MR and can corroborate a patient’s epileptogenic history as seen in MRI. D3.1.1.4 Vessel Occlusion D3.1.1.4.1 Arterial Occlusion As the brain is extremely energy and oxygen dependent, any decrease in cerebral blood flow will affect the brain. The effect—or rather the damage—can vary greatly depending on the location, the extent, and the duration of blood flow decrease. If the cerebral oxygen level is depleted due to a systemic disorder such as a cardiac arrest or suffocation, the whole brain will be affected, with certain especially oxygendependent regions being damaged first. This generalized hypoxia will result in a cerebral edema and eventually death. Localized cerebral ischemias are generally due to an occlusion of a cervical or cerebral artery. These occlusions may be due to thrombemboli and atherosclerotic plaque ruptures. Arteriosclerotic vessel narrowing of one artery is rarely a cause for cerebral ischemia as this process takes weeks to years and can be compensated by the arterial shunts of the circle of Willis. The effects of ischemias may be reversible, as in transitory ischemic attacks, or may lead to infarction, clinically referred to as stroke or insult of the brain originally
© 2009 by Taylor & Francis Group, LLC
225
FIGURE D3.1.1.5 Right mesial temporal sclerosis with T2hyperintense signal and minor volume decrease compared with the opposite site.
supplied by the occluded artery. At autopsy, the diagnosis of cerebral insult may be difficult. Small or hyperacute insults can easily be overlooked if no histological sampling and examination is performed. MR imaging is better suited than autopsy concerning acute cerebral insult, as it offers an early and reliable diagnostic tool for the detection of hypoxic brain lesions [10,11] (Figure D3.1.1.6). In the T2-weighted MR images, the hypoxic tissues can appear as hyperintense areas after 2–8 hours. The gray–white differentiation is often reduced, and signs of brain swelling can be present [10,11]. In native CT, a hyperdense blood vessel might be seen, and often the differentiation between gray and white matter is reduced. As a fact in postmortem examination and due to the suspended circulation, there is a lack of performing perfusion or contrast-enhanced imaging. (see also Chapter D3.3, “Forensic Neuroimaging.”) These techniques are applied in daily clinical practice when it comes to diagnosing patients with the suspicion of cerebral infarction or vessel disease. There are, however, attempts to develop techniques for a postmortem application of contrast media [12–15], and the results of these studies are promising. Future forensic-radiological research will define the value and possible applications for these technologies. D3.1.1.4.2 Thrombosis of the Venous Sinuses The most important venous complication is thrombosis of the venous sinuses. Three different thromboses are known: r Thrombosis of the superior sagittal sinus r Thrombosis of the cortical veins r Thrombosis of the central cerebral veins
226
The Virtopsy Approach
A
B
C
FIGURE D3.1.1.6 (A) Acute hemorrhage in a 92-year-old patient who fell to the floor. Massive intraparenchymal bleeding with rupture to the ventricles (CT examination). Traumatic intraparenchymal bleeding in CT (B) and MRI T2 sequence (C).
All of these disorders are rare but can occur spontaneously or as a complication of other diseases, like infectious origins (e.g., mastoiditis, tonsillitis, sinusitis, tuberculosis) or hematological sources (e.g., coagulation disorders, leukemia). Headache, hemiparesis, papillary edema, insults, and a lowered awareness are symptoms in thrombosis of the superior sagittal sinus. Aphasia can occur but is rare. Focal symptoms of the cortical venues thrombosis are hemiparesis, often combined with epileptical insults. With lumbar punction, a high intraspinal pressure of the spinal fluid is found, mostly with no other symptoms [16–18]. The main problems of postmortem CT and MRI concerning the detection of thromboses of the venous sinuses are the arrest of blood circulation in the dead. In the near future, it might be possible to detect these thromboses due to the research in postmortem angiography (see Chapter D6, “Angiography”). D3.1.1.5 Hemorrhagic Lesions
Whatever the cause of the nontraumatic subarachnoid hemorrhage, as soon as blood enters the subarachnoid space it causes a mild inflammatory reaction in the meninges as well as vasospasms. These vasospasms can potentially cause an insult. Fibrosis subsequently develops in many cases. Fibrosis of the pia matter develops in about 10 days. Since slight fibrosis of the pial and arachnoid membranes may be present as a “normal” aspect of the membranes, especially with advancing age, interpretation of minimal fibrosis is difficult (Figure D3.1.1.7) [16,19,20]. D3.1.1.5.2 Intracerebral Hemorrhage Intracerebral hemorrhage is characterized clinically by an abrupt onset of a very intense headache and leads rapidly to coma. Intracerebral hemorrhages occur more often in males than in females and show a higher incidence in persons of African origin, probably due to the greater incidence of hypertension [21,22].
Nontraumatic intracranial hemorrhages concern the brain coverings such as in subarachnoid and subdural hemorrhages (and very rarely the epidural space) as well as intracerebral hemorrhages. D3.1.1.5.1 Nontraumatic Subarachnoid Hemorrhage The second most common cause of sudden unexpected death due to natural disease of the brain is nontraumatic subarachnoid hemorrhage. Early in the 20th century, spontaneous subarachnoid hemorrhage was considered a disease entity in itself. With the advancement of medical knowledge came the realization that it was a syndrome with multiple causes. Aneurysms are the most common cause of subarachnoid hemorrhage, followed by intracerebral hemorrhages and, to a lesser degree, rupture of arteriovenous malformations. Arteriovenous malformations probably account for only a few percent of nontraumatic subarachnoid hemorrhages. They tend to cluster in the early decades of life, though they can be found at any age. Uncommon causes of nontraumatic subarachnoid hemorrhage would be blood dyscrasias, endocarditis with embolic phenomenon, overuse of anticoagulants, tumors (primary or metastatic), and sickle cell disease.
© 2009 by Taylor & Francis Group, LLC
FIGURE D3.1.1.7 Spontaneous intraventricular bleeding with blood in the subarachoindal space.
Incident-Specific Cases
People of older age have intracerebral hemorrhages more often than people in younger age groups. Physical activities provoke the hemorrhage in people with a preexisting hypertension. Unlike subarachnoid hemorrhages, recurrence of the hemorrhage is uncommon. Causes of nontraumatic intracerebral hemorrhages are as follows: 1. Hypertension: Hyaline degeneration of the medial tunica of the cerebral arteries can evolve under the influence of hypertension. The predilection place is the lenticulostriate artery. If this vessel ruptures, a bleeding into the internal capsule and the basal ganglia occurs. 2. Anticoagulation: High bleeding tendency causes are mainly found in anticoagulation, thrombopenia, and hemophilia, often in combination with light traumas. 3. A berry or saccular aneurysm: If the aneurysm ruptures, a subarachnoid or intracerebral hemorrhage can occur. 4. Arteriovenous malformations (AVMs): Mostly these occur in malformations in or on the cortex, which can cause not only bleedings but also seizures. 5. Bleeding of a hypervascular tumor (e.g., Glioblastoma multiforme) and metastasis. Depending on the site of the bleeding, different symptoms can occur, but the main complaints are complete hemisyndrome, paresis, and loss of view. Sensory deficit is usually greater than motor weakness. As the hemorrhage often continues to bleed, awareness decreases over the hours and results in coma. In hemorrhages into the pontine area, loss of consciousness is almost immediate. Asymmetrical swelling of the brain can occur in intracerebral hemorrhage, where the swollen hemisphere contains the hemorrhage. Subarachnoid hemorrhage can be present when the bleeding reaches the cortex of the ventricles. The brain tissue surrounding the hemorrhage is swollen and edematous. Hemorrhage can cause compression and distortion of the midbrain, which results in death, or it can bleed into the ventricles. Although a ruptured saccular aneurysm can be seen as a natural cause of death, the history must be examined carefully. The main aspect of the examination is the exclusion of previous trauma, which may be responsible for the bleeding resulting in death [17–19,23,24]. The differential diagnosis of hypertensive intracerebral bleeding versus trauma-induced intracerebral hemorrhage is important in forensic examinations. Acute hypertonic intracerebral hemorrhage is often extensive, is usually localized in the basal ganglia and the central brain regions (capsula interna), and is less frequent in the pons and cerebellum, whereas traumatic bleeding is rather peripheral and subcortical with accompanying fracture respectively the impact site. Usually hypertonic bleeding can be diagnosed by both
© 2009 by Taylor & Francis Group, LLC
227
CT (hyperdense area) and MRI (sequence and time-dependent) without difficulties in nontraumatic cases; a solitary hematoma with a local edema and no other injuries is seen. Hemosiderin pigments in chronic bleeding shows hypointense borders in T1 and T2 sequences. In case of a nontraumatic genesis, the hematoma will be solitary, without other accompanying traumatic injuries but with a local surrounding edema. D3.1.1.6 Vascular Malformations The most common lethal vascular malformations are composed of aneurysms and arteriovenous malformations. Angiomas, which rarely rupture when, they usually rise to discreet symptoms, can present as foci for epileptic seizures. D3.1.1.6.1 Aneurysms Intravascular pressure exploits weakness in arterial walls and causes aneurysms. Among the many causes of aneurysms are the following: r Developmental defects, which give rise to berry, saccular, or media sclerotic aneurysms r Atherosclerosis and hypertension r Trauma r On rare occasions trauma may give rise to dissecting aneurysms (see Section D3.3, Forensic Neuroimaging”) r Bacterial infections D3.1.1.6.1.1 Berry (Saccular) Aneurysms Usually, these aneurysms are the result of a congenital defect during embryonic development at the time an artery separates itself to receive a Y-shaped configuration. The main vessel and the two branches may fail to interdigitate adequately across the Y-notch, which creates a point of muscular weakness. The main bloodstream exerts pressure at the Y-notch. Due to this, most of the berry aneurysms are located at branch points in the Willis circle and the carotid system. In one to five cases, multiple berries are found at autopsy. Time and trauma will expose this defect. A berry aneurysm evolves, with its wall being formed precariously of the adventitia. Berry aneurysms increase in frequency with age; they are rare in children but are found in 25% of persons older than 55. These aneurysms may enlarge, even to such an extent that they can compress the brain and the nerves and result in palsies or other neurological symptoms. Hypertension and cigarette smoking are predisposing factors of atherosclerosis, leading to focal destruction and weakening of the vessels walls. A rupture of these aneurysms gives rise to a subarachnoid hemorrhage, which may also bleed into the subdural space and even into the brain. This can occur at any age but has the highest frequency for people between 30 and 50 years of age. Sixty percent of such affected persons died immediately after rupture. Of the survivors of the initial hemorrhage,
228
more than half die less than 24 h after admission to the emergency room. Patients who survive usually have a progressive decline in consciousness after 3 to 4 days, an effect that is the result of a generalized arterial spasm and consequent cerebral ischemia and infarction. Prognosis is worse in patients who rebleed after the initial hemorrhage. Berry aneurysm rupture is the most common cause of nontraumatic life-threatening subarachnoid hemorrhage, with a mortality rate of 35% during the initial bleeding. A rupture of a berry aneurysm can also result in intracerebral and intraventricular hemorrhage in 33% of patients, although they generally bleed into the subarachnoid space. Upon rupture of a berry aneurysm, the patient usually complains of a sudden excruciating headache, followed by nausea, vomiting, and then loss of consciousness. The blood in the subarachoidal space triggers a generalized vasospasm of the brain vessels, which results in ischemic lesions to the brain. In cases of headaches that have lasted for weeks, a minor leakage in the vessels may cause the discomfort and may rupture with fatal results. In autopsy, the largest amount of blood is on the ventral side of the brain, with a lesser amount on the dorsal and lateral side. A large pool of blood on the ventral surface of the brain often makes it difficult to locate the aneurysm during autopsy if the brain is not examined when fresh [1,17,19,23]. In cases of aneurysmatic subarachnoid hemorrhage, CT will reveal a hyperintensity in the basal subarachnoid spaces, which can circulate in the entire subarachnoidal space. MRI is a little less sensitive for the detection of this finding, as the hemorrhage often appears isointense to the brain parenchyma. (See also Chapter D3.3, Table 3.3.1.) Concerning the detection of the aneurysm that caused the subarachnoid hemorrhage, autopsy is superior to forensic imaging due to the lack of adequate postmortem angiographic methods (see also Section D3.1.1.3, “Epileptic Disorders”). More research is necessary in postmortem angiography and is being performed by the Virtopsy Research Group at the University of Berne in Switzerland (see Chapter D6, “Angiography”). D3.1.1.6.1.2 Arteriovenous Malformations AVMs are mostly found in the parietal cortex of the brain, appearing as a wedge of arteries and veins extending into subcortical white matter. Deep AVMs may lie in the white matter, basal ganglia, thalamus, or brainstem. These are complex tangles of abnormal arteries and veins linked by one or more fistulas. They lack a capillary bed, and the small arteries have a deficient muscularis [16,25] (Figure D3.1.1.8). A small percentage of the cases of nontraumatic subarachnoid or parenchymal hemorrhage are due to bleeding from an AVM. Most AVMs derive part of their blood supply from at least one branch of the middle cerebral artery. There may be severe bleeding from these lesions into the subarachnoid space or into the substance of the brain, presenting as a massive intracerebral hemorrhage. It is estimated that 0.1% of the population has AVMs, with 12% of these becoming
© 2009 by Taylor & Francis Group, LLC
The Virtopsy Approach
FIGURE D3.1.1.8 Postgadolinium T1 fat sat sequence: parietal AVM with contrast agent-rich capillaries.
symptomatic [16]. Sturge-Weber syndrome is characterized by multiple AVMs of the cerebral hemispheres associated with vascular nevi of the face or neck combined with epilepsy. The most common clinical presentations are intracranial hemorrhage (30%–82%), seizures (16%–53%); headache, and focal neurological deficits [16]. Two percent of all strokes are due to AVMs. The death rate from hemorrhage is 10% to 15%. D3.1.1.7 Inflammatory Processes To date, experiences concerning forensic postmortem imaging of inflammatory processes inside the cranium are rare. Based on the aforementioned clinical radiologic experience, it is likely that postmortem CT and MRI have a certain potential of depicting inflammatory changes before autopsy. This could help the forensic examiner to plan and apply special examination techniques (e.g., liquor puncture, tissue sampling) before entirely opening the cranial cavity and therefore could help prevent autopsy personnel from being contaminated with infectious material. In the following sections, the typical radiological signs of native CT and MRI are described for the main forms of intracranial infection. The application of contrast media is used for clinical examination in daily radiological routine; however, these techniques to date cannot be applied in postmortem situations, but research is done in postmortem angiography. Another safety advantage in safety of imaging before autopsy is the possibility for biopsy for (neuro-) pathological, histological, and or microbiological examination. Only small punctures are necessary for the biopsy needle and
Incident-Specific Cases
229
A
B
C
D
E
FIGURE D3.1.1.9 Suicide with handgun. Notice the (malignant) brain tumor (red circles) in (A), (B), and (C). In (D) the pneumocranium is visible (blue arrow). (E) shows the fractures (red arrows) of the skull caused by the projectile (See also Chapter D3.5, “Gunshot”).
can be performed by one experienced examiner. This way, personnel contamination risks and the examination area are reduced to a minimum (see Chapter D5, “Biopsy”). D3.1.1.8 Tumorous Lesions The brain can also be the affected site due to primary or secondary tumors. A sudden death due to an undiagnosed primary brain tumor is rare. In children, the second most common cause of death is primary brain tumors, with hematological tumors being the most common cause in this group [26]. Due to the growing tumor, more room is needed, the pressure on the brain tissue rises, and symptoms—and therefore medical diagnosis—occur relatively early. This is why primary brain tumors rarely cause a sudden death and therefore seldom necessitate a forensic examination. In a study of 10,995 consecutive medicolegal autopsies in Dallas, Texas, Di Maio et al. found 19 sudden, unexpected deaths due to primary intracranial neoplasms, an incidence of 0.17%. In another study of 17,404 autopsies performed at the Brooklyn Office of the Medical Examiner, DiMaio et al. found an incidence of 0.16% of sudden, unexpected deaths due to primary intracranial neoplasm [27–30]. To give a general pathologic overview of the main forms of intracranial tumors and their radiological appearance is beyond the scope of this chapter. In addition, in contrast to neuropathology, tumor alterations are, as shown already, of minor relevance in forensic casework, although in some cases the tumor can be a reason for suicide (Figure D3.1.1.9).
© 2009 by Taylor & Francis Group, LLC
Neuroradiological literature gives information about the tumor characteristics [10]. For example, in CT a malignant tumor is isodense and in MRI is also isointense. Perifocal edema in CT is hypodense, in an MR T2 sequence is hypointense, and in T2 is hyperintense. Tumoral bleeding is hyperdense in CT and is hyperintense in T1 sequences. D3.1.1.9 References 1. DiMaio, V.J. and D. DiMaio. 2001. Deaths due to natural disease. In Forensic pathology, 2d ed. CRC Press, London. 2. Reimann, W. and O. Prokop. 1985. Vademecum Gerichtsmedizin, 4th ed. [Vademecum Forensic Medicine, authors’ translation], VEB Verlag Volk und Gesundheit, Berlin. 3. Aghayev, E., K. Yen, M. Sonnenschein, et al. 2004. Virtopsy post-mortem multi-slice computed tomography (MSCT) and magnetic resonance imaging (MRI) demonstrating descending tonsillar herniation: Comparison to clinical studies. Neuroradiology 46(7):559–64. 4. Blisard, K.S. and P.J. McFreely. 1998. The spectrum of neuropathologic findings in deaths associated with seizure disorders. J Forensic Sci 33(4):910–14. 5. Coyle, H.P., N. Baker-Brian, and S.W. Brown. 1994. Coroners’ autopsy reporting of sudden unexplained death in epilepsy (SUDEP) in the UK. Seizure 3–4:247–54. 6. Leestma, J.E., J.R. Hughes, S.S. Teas, M.B. Kalelkar. 1985. Sudden epilepsy deaths and the forensic pathologist. Am J Forensic Med 6(3):215–18. 7. Terrence, C.F., G.R. Rao, and J.A. Perper. 1981. Neurogenic pulmonary oedema in unexpected, unexplained death of epileptic patients. Ann Neurol 9(5):458–64.
230
8. Falconer, B. and J. Rajs. 1976. Post-mortem findings of cardiac lesions in epileptics: a preliminary report. Forensic Sci 8(1):63–71. 9. Freytag, E. 1966. Fatal rupture of intracranial aneurysms. Arch Pathol 81:418–24. 10. Randall, B.B., M.F. Fierro, and R.C. Froede. 1998. Practice guideline for forensic pathology. Arch Pathol Lab Med 122:1056–64. 11. Yen, K., J. Weis, R. Kreis, et al. 2006. Line-scan diffusion tensor imaging of the posttraumatic brain stem: changes with neuropathologic correlation. AJNR Am J Neuroradiol 27(1):70–73. 12. Jackowski, C., M. Sonnenschein, M.J. Thali, et al. 2005. Virtopsy: postmortem minimally invasive angiography using cross section techniques—implementation and preliminary results. J Forensic Sci 50(5):1175–78. 13. Jackowski, C., S. Bolliger, E. Aghayev, et al. 2006. Reduction of post-mortem angiography-induced tissue oedema by using polyethylene glycol as a contrast agent dissolver. J Forensic Sci 51(5):1134–37. 14. Grabherr, S., V. Djonov, A. Friess, et al. 2006. Postmortem angiography after vascular perfusion with diesel oil and a lipophilic contrast agent. AJR Am J Roentgenol 187(5):W515–W523. 15. Grabherr, S., V. Djonov, K. Yen, M.J. Thali, and R. Dirnhofer. 2007. Postmortem angiography: review of former and current methods. AJR Am J Roentgenol 188(3): 832–38. 16. Leestma, J.E. 1988. Forensic Neuropathology. New York, Raven Press. 17. Kuks, J.B.M., J.W. Snoek, and H.J.G. Oosterhuis. 2003. Klinische Neurologie, 15th ed. [Clinical neurology, authors’ translation]. Houten / Diegem, Bohn Stafleu Van Loghum. 18. Hijdra, A., P.J. Koudstaal, and R.A.C. Roos. 2003. Neurologie, 3d ed. [Neurology, authors’ translation]. Reed Business Information, Maarssen. 19. Davis, R.L. and D.M. Robertson. 1991. Textbook of Neuropathology, 2d ed. Williams and Wilkins, Baltimore. 20. Unterharnscheidt, F. 1992. Spezielle pathologische anatomie. Bd. 13 Pathologie des Nervensystems VII [Special pathological anatomy. Bd. 13 pathology of the nervous system VII, authors’ translation]. Springer Verlag, Berlin. 21. Hajat, C., R. Dundas, J.A. Stewart, et al. 2001. Cerebrovascular risk factors and stroke subtypes: differences between ethnic groups. Stroke 32(1):37–42. 22. Labovitz, D.L., A. Halim, B. Bodem-Albala, W.A. Hauser, and R.L. Sacco. 2005. The incidence of deep and lobar intracerebral hemorrhage in whites, blacks and Hispanics. Neurology 65(4):518–22. 23. Rubin, E. and J.L. Farber. 1994. Pathology, 2d ed. J.B. Lippincott, Philadelphia. 24. Martin J.H. 1989. Neuroanatomy. Appleton & Lange, East Norwalk, CT. 25. Karhunen, P.J., A. Penttilä, and T. Erkinjuntti. 1990. Arteriovenous malformation of the brain: imaging by postmortem angiography. Forensic Sci Int 48:9–19. 26. Nelson, J., J.L. Frost, and S.S. Schochet Jr. 1987. Sudden, unexpected death in a 5-year-old boy with unusual primary neoplasm. Ganglioglioma of the medulla. Am J Forensic Med Pathol 8:148–52. 27. Abu Al Ragheb, S.Y., K.J. Koussous, and S.S. Amr. 1986. Intracranial neoplasms associated with sudden death: a report of seven cases and a review of the literature. Med Sci Law 26(4):270–72.
© 2009 by Taylor & Francis Group, LLC
The Virtopsy Approach
28. Di Maio, T.M. and D.J. Di Maio. 1974. Sudden death due to colloid cysts of the third ventricle. N Y State J Med 74(10):1832–34. 29. Di Maio, S.M., V.J. Di Maio, and J.B. Kirkpatrick.1980. Sudden, unexpected deaths due to primary intracranial neoplasms. Am J Forensic Med Pathol 1:29-45. 30. Gleckman, A.M. and T.W. Smith. 1998. Sudden unexpected death from primary posterior fossa tumors. Am J Forensic Med Pathol 19(4):303–08.
D3.1.2 CARDIAC PATHOLOGY Christian Jackowski and Michael Thali D3.1.2.1 Background Magnetic resonance imaging (MRI) and computed tomography (CT) are well-established tools in noninvasive cardiac diagnostic with a fast development in the recent years [1–3]. The clinical radiological examination is commonly a diagnosis of function and is mostly based on contrast media flow in the ventricles/coronary arteries, motion, and kinetics. Cardiac CT gained clinical importance in estimating the calcium score of the coronaries [4], reached impressing results in 3D coronary angiographies [5], and allows motions analysis and thereby the assessment of kinetic disorders [6]. Although late enhancements of the myocardium after contrast agent administration can also be visualized in CT as a diagnostic hint for ischemic myocardium [7], the detailed morphological assessment of myocardial alterations requires MRI. This is what generally parallels postmortem imaging of the human heart. However, postmortem imaging of the heart is a nonfunctional but morphological assessment. Not having to overcome cardiac motion while generating images of the organ, the quality of the images is superior to clinical experiences. Instead of motion artifacts and blood-flow phenomena, postmortem alterations such as putrefaction and postmortem reactions of the blood (internal livores, postmortem clotting), as well as the variability of the body temperature of the corpse, all influence the imaging process. The first studies in 2003 on isolated human autopsy hearts [8] and entire corpses [9,10] show that these imaging techniques have great potential in the postmortem diagnostic of pathologies of the heart, and these studies also demonstrate the influence of the postmortem reactions on the imaging. The first study exemplarily demonstrating the overall potential of postmortem cardiac imaging arose in 2005 [11]. Recent research efforts concentrate on the validation of experiences that have been acquired in single or a few cases using larger study populations. This work is still done with the methods of the first studies such as comparing imaging appearances to the findings in a traditional autopsy as the gold standard of postmortem investigation during the last centuries. D3.1.2.2 How to Get Cardiac Cross-Sectional Images Although CT and MRI can acquire the whole-volume data of the heart and thereby a 3D reconstruction of the heart
Incident-Specific Cases
231
B
Short axis RV
LV
Plane for vertical long axis
A
L
(b) RV
(d)
LV
L Base
A
LA L
LV
Apex A Apex
(c) C
D L
L
(d)
Base LA Plane for horizontal long axis
RA RV
FIGURE D3.1.2.1 Sketch showing defined cardiac imaging planes and their 3D combinations. RV, right ventricle; LV, left ventricle.
is possible and in many cases very useful for visualization, however, one will start to look at the 2D images first. The routinely generated axial (transversal) images of the thorax are of inferior value as they show very oblique sections through the heart that depend on the individual position of the organ within the thorax. To compare different studies and examinations, it is necessary to define cardiac sections that are used to make the diagnosis. There are three main 90n-oriented planes: The terms for the cardiac planes used in all imaging modalities are short axis, vertical long axis, and horizontal long axis (Figure D3.1.2.1). These correspond to the short-axis, two-chamber, and four-chamber planes traditionally used in 2D echocardiography. One has to define the planes to scan in MRI at the time of scanning. Starting with a transversal thoracic image showing the heart, the next plane (vertical long axis) will be located through the left chamber parallel to the septum (Figure D3.1.2.2). On the resulting vertical long-axis images, one has to locate the next plane (horizontal long axis) longwise through the left ventricle. Having located the horizontal long axis, the next plane (short axis) is situated parallel to the base of both ventricles. Alternatively, one can choose the vertical long-axis images to define the short-axis planes as parallel to the base of the left ventricle [12]. D3.1.2.3 Postmortem Alterations D3.1.2.3.1 Postmortem Clot and Sedimentation of Blood Components (Internal Livores) The formation of livores contains much forensic-relevant information, such as the position of the corpse during the postmortem period. Its accruement is based on the sedimentation of the cellular components (prevailing erythrocytes) of the blood after the stagnation of circulation. This phenomenon is
© 2009 by Taylor & Francis Group, LLC
RV
LV Li
LV
L
FIGURE D3.1.2.2 How to get cardiac imaging planes starting with an axial thoracic image? After searching for axial thoracic images that show a section through the heart (A) the next plane is positioned through the left ventricle parallel to the interventricular septum (B). This should result in a vertical long-axis image set sketched in (B). The following horizontal long-axis image set (C) is located on the vertical long-axis image (B) as a longwise section through the atrium and the ventricle. Within the horizontal longaxis images (C) the new located planes parallel to the base (plane of the valves) will result in short-axis images (D). Alternatively, one may try to locate short-axis images on the vertical long-axis image set as demonstrated in (B). RV, right ventricle; LV, left ventricle; L, lungs; RA, right atrium; LA, left atrium; A, aorta; Li, liver.
impressively visible in T2-weighted axial images of the thorax showing the cardiac cavities or the main thoracic vessels [13]. The upper serum layer appears as very bright in contrast to the sedimented erythrocytes that are of distinctively lesser signal (Figure D3.1.2.3). At external examination these sedimentation phenomena appear as livores on the skin. Otherwise, they are also detectable within internal organs such as the lungs, kidney, liver, intestine, and heart. Within the lungs they are the most distinctively visible in CT [10]. In the heart they can appear in T2-weighted images as hypointensive regions in the lower areas of the myocardial wall of the left ventricle (Figure D3.1.2.3) and show dense-filled intramyocardial vessels in histological examination. They should not be misinterpreted as vital myocardial alteration by comparing this finding to the position of further livores. Their appearance as well as of further livores is dependent on the remaining blood volume after the dying process and the position of the corpse after death. Another postmortem alteration of the blood visible in MR images is postmortem coagulation, known as postmortem
232
The Virtopsy Approach
A
C
B
D
FIGURE D3.1.2.3 Physiological postmortem findings. Sedimentation of the cellular blood components results in a layering of the blood within the cardiac cavities (dashed arrows in (A)). Sedimentation within the tissue of the organ (called internal livores) can cause lower signal in T2-weighted images in depending regions of the myocardium (arrows in (A)). Histology shows dense erythrocyte filled intramycardial vessels in these areas (arrows in (B)). Postmortem clots can fill almost the entire cardiac cavities as seen in (C) (arrows, arrowhead indicated a sedimentation level within the left ventricle). Postmortem clots easily slip out of the ventricle at autopsy, when the heart is opened (D).
clotting. It is frequently seen in cases with prolonged agonal periods. Depending on the prefinal function of the vital clotting system, the postmortem clotting process occurs rapidly and encloses the cellular components of the blood (mostly erythrocytes), and a cruor coagulation (red) accrues. It occurs slowly and the erythrocytes already sedimented out and are not enclosed and a lighter (white) clot accrues. The process is not triggered by a defect of the intima so that the clots do not adhere to the endocardial wall. Postmortem clots present in T2-weighted MR images as some bizarre structures in the center of both ventricles without any direct contact to the endocardium or nearly filling the ventricle and therefore touching the endocardium at the lower site as a secondary effect (Figure D3.1.2.3). During autopsy, the postmortem clots easily slip out of the ventricle after opening it (Figure D3.1.2.3). Depending on the content of cellular components, it is hyperintense, isointense, or hypointense in postmortem MRI [13]. In contrast to vital intracardial thrombi, postmortem clots do not
© 2009 by Taylor & Francis Group, LLC
especially adhere to myocardial infarction areas or hypokinetic regions and do not show any hemosiderin-induced loss of signal [14] as mature vital thrombi can do. The occurrence of distended postmortem clots is a forensic hint to a prolonged agony and can also be acquired by postmortem MRI. D3.1.2.3.2 Cardiac Rigor Mortis Comparably to the rigor mortis within the skeletal musculature, the heart remains in a systolic situation when adenosine-tri-phosphate (ATP) reserves of the myocardium are exhausted. Being a ring muscle, this leads to an increase of intramyocardial pressure from subepicardial to subendocardial regions. Thereby interstitial fluids arrange in less pressured subepicardial regions and cause a signal decline in MR images from subepicardial to subendocardial when the rigor mortis is at its maximum (postmortem interval [PMI] 24–48 h) or the myocardium is hypertrophic.
Incident-Specific Cases
D3.1.2.3.3 Putrefaction Putrefaction is a postmortem alteration that complicates diagnostics in traditional autopsy as well as in postmortem imaging. Therefore, it is of utmost importance to know its radiological appearance to be able to differentiate between further pathologic alterations. In CT, putrefaction is indirectly reflected by its gas formation. This gas contains CH4, NH3, H2S, CO2, H2,N2, cadaverin, and putrescin and is a result
A
233
of the proteo-katabolic process by anaerobic bacteria such as proteus- and coli- species and bacillus subtilis. As a component of the physiological flora of the intestine, the putrefaction bacteria reach the organs mainly via the vessels after death. Therefore, in the heart there are at first small gas bubbles at the endocardial site of the myocardium seen in CT. Later, a confluence of the small bubbles to bigger intraventricular ones situated in the upper lying parts of both ventricles or completely filling the cardiac caves arises (Figure D3.1.2.4). The gas is also seen in MR images as bubbles with low signal. Later on, the loss of the myocardial structure can be seen in T2-weighted MR images, beginning on the endocardial site of the ventricles (Figure D3.1.2.4). The postmortem formation of intracardial gas should not be confounded with air embolism (Figure D3.1.2.5). This can be discerned by
A
B
B
FIGURE D3.1.2.4 Cardiac putrefaction (A) by MSCT, putrefaction gas-filled ventricles (arrows) and small subendocardial gas bubbles (dashed arrow) on the axial thoracic view; note the structure accentuation of thoracic muscles and visceral organs caused by interstitial and intraluminal putrefaction gas inside and outside the rib cage as well as pleural putrefaction fluid; and (B) by MRI, there is myocardial signal loss, mainly subendocardially (arrows on a T2-weighted axial image). Putrefaction gas also causes susceptibility artifacts and induces loss of signal within both ventricles and the pericardium and pleura. In addition, there is bilateral putrefaction fluid in the dependent pleural and pericardial spaces.
© 2009 by Taylor & Francis Group, LLC
FIGURE D3.1.2.5 Right-ventricular venous air embolism with complete filled ventricle (arrow in (A)). Note the venous air distribution within the adjacent vessel system in the magnification given in (B). Air-filled epicardial veins (white arrows) next to the coronaries (yellow arrows) and air-filled internal thoracic veins (white dashed arrows) next to the internal thoracic artery (yellow dashed arrow).
234
comparing the degree of putrefaction in further organs or by the appearance of gas in both ventricles in absence of a septal defect. In nondistinctive cases, a puncture of both ventricles at autopsy or percutaneous [15] followed by an analysis of the gas provides clarification by showing comparable consistence of the gas of both ventricles. A significant difference between both ventricles with a higher proportion of putrefaction gas in the left ventricle may imply relevant venous air embolism despite beginning putrefaction.
The Virtopsy Approach
SP a
i b
j c
D3.1.2.4 Pathological Findings
d
D3.1.2.4.1 Gas Embolism
e
Gas embolism is a frequent forensic finding in cases of injury to the head and neck as a venous air embolism [16–19]. Depending on the air volume embolized, it may be the primary cause of death, an assisting or competing cause of death, or an accessory finding [20]. Visualization and quantification at the autopsy room have been a great challenge through the past centuries and could not lead to satisfactory results. Opening the heart under water as well as traditional chest radiographs [16] failed in quantification. Elaborate methods using an aspirometer [21,22] could not reach widespread application. CT can impressively demonstrate the intrathoracic air and its distribution and can quantify the air volume that is within the vascular system or the heart (Figure D3.1.2.5, Figure D3.1.2.6, Figure D3.1.2.7, and Figure D3.1.2.8). D3.1.2.4.2 Calcification The atherosclerosis of the coronary arteries and its complications are a common cause of death. Therefore, it is important
FIGURE D3.1.2.6 Complete air-caused occlusion of the pulmonary artery system in a case of venous air embolism (arrow). Note that it is not necessary to quantify the embolized air volume anymore, as it is clearly demonstrated that the air amount was able to completely occlude the artery and thereby caused the death.
© 2009 by Taylor & Francis Group, LLC
f
k
g
l h
FIGURE D3.1.2.7 AP view on a 3D air reconstruction of a typical air distribution within the vascular system in a case of a venous air embolism caused by a gunshot wound to the head: (a) Right external jugular vein. (b) Right internal jugular vein. (c) Right subclavian vein. (d) Superior vena cava. (e) Right upper lobe artery. (f) Right middle lobe artery. (g) Right lower lobe artery. (h) Small cardiac vein. (i) Left internal jugular vein. (j) Left subclavian vein. (k) Left pulmonary artery. (l) Right ventricle.
to have a diagnostic tool that can show its spread on the coronary arteries. Via the deposition of calcium hydroxyapatite (calcification) to the intima, the wall of the vessels shows an increased radiological density [23]. This leads to 130 – 700 Hounsfield units for calcification including voxels. These depositions are easily seen in multislice CT (MSCT) (Figure D3.1.2.9). According to the spread of the calcification, one may create a 3D reconstruction of almost the entire coronary artery system using a volume-rendering reconstruction protocol (Figure D3.1.2.9) or may quantify the deposition of calcium by scoring the peak densities of identified deposits [24]. Multiple studies have shown the correlation between the deposition of calcium and the narrowing of the artery lumen [25,26]. However, these are only statistical coherences and do not predict the actual grade of the certain stenosis. In MR images, the calcifications cause a reduction or loss of signal indicating the coronary lesion. The knowledge of the actual coronary lumen is far more important for a diagnosis such as acute myocardial infarction. It is therefore still necessary to have a minimal invasive method for visualization of the coronary arteries (see also Chapter D6, “Angiography,” Figure D6.2.16). For isolated hearts, for decades there have been techniques to visualize the coronary artery system using contrast enhanced x-ray or
Incident-Specific Cases
235
A
a
B
b
c
d
c
a
b
FIGURE D3.1.2.8 Cardiac air distribution in a case of a venous air embolism after a gunshot wound to the head and a patent foramen ovale (A). Note that the air can be visualized within the rightventricle (white arrow), right atrium (white dashed arrow) as well as within the left atrium (yellow arrow). Magnification (B) reveals the gaping of the septum primum from the septum secundum as the way the air reached the left-sided cardiac cavities (white arrow).
CT [27]. Minimally invasively applied on the entire human corpse, angiographic methods using radiopaque contrast agents and CT have proven to be feasible to display also the coronaries in sufficient quality to assess narrowing or occlusions [28,29] (Figure D3.1.2.10). Further locations for the deposition of calcium are the cardiac valves (Figure D3.1.2.11). Particularly the stenosis of the aortic valve is caused by calcification. Studies have shown that there is a significant correlation between the verified amount of calcium by CT and the degree of the stenosis [30]. So MSCT provides a possibility of giving an indirect impression of the degree of the stenosis. In combination with documented left-ventricular (LV) hypertrophy and acute edema of the lungs, it can make the diagnosis of an acute LV insufficiency as cause of death caused by stenosis of the aortic valve likely. Further common cardiac locations of calcifications are the papillary muscles. These are a phenomenon of the senescent
© 2009 by Taylor & Francis Group, LLC
FIGURE D3.1.2.9 Calcification of the coronary artery system by postmortem axial CT (upper image); the right coronary artery (a), the circumflex coronary artery (b), and left anterior descending coronary artery (c) are calcified, as well as scars of the pleura (d). 3D volume rendering in severe atherosclerosis of coronary arteries (lower image). On this craniocaudal view on a 3D volume rendering model of the heart and parts of the thoracic spine extended calcification of the left descending coronary artery (a), circumflex coronary artery (b), and right coronary artery (c) is displayed. Note also a calcified leaflet of the aortic valve.
heart as well as a result of local ischemia [31–33]. They are of less forensic importance, except for the lethal acute insufficiency of the mitral valve arising due to rupture of a mitral chordae tendineae or the papillary muscle. In MSCT (Figure D3.1.2.12) they appear as intraventriculiar spots of higher radiological density easily assignable to the papillary muscle system [8]. D3.1.2.4.3 Endocarditis Bacterial infections of the endocardium preferentially affect the cardiac valves. These valvular vegetations are a feared complication of endocarditis because they can lead to
236
The Virtopsy Approach
A
B
FIGURE D3.1.2.11 Calcification of the aortic valve in postmortem MSCT. Two radiopaque spots of calcification (yellow arrows) that are located at leaflets of the aortic valve are shown. Note also the air embolism in the right ventricle and a significant pericardial thickening by effusion.
a b
FIGURE D3.1.2.10 Curved image of a minimal invasive postmortem CT angiography of the supravalvular aorta (dashed arrow) with the left coronar ostium (A). Note the calcified plaque (white arrow) and the sufficient occluding leaflets of the aortic valve (yellow arrows). A curved image along the left anterior descending coronary artery (B) demonstrates a calcified plaque (arrow) but excludes significant narrowing.
irreversible stenosis or insufficiency of the affected valve or can cause peripheral emboli and cerebral stroke. The diagnosis of valvular vegetations (Figure D3.1.2.13) is vitally made by transthoracic/transesophagial echocardiography. Postmortem MRI clearly shows vegetations of the cardiac valves without motion artifacts and with an adequate contrast (Figure D3.1.2.13). Further postmortem microbiological investigations will support this diagnosis. D3.1.2.4.4 Ischemia Although cardiac and pulmonary motion artifacts aggravate diagnostics, clinical radiology has the advantage of contrast
© 2009 by Taylor & Francis Group, LLC
d
c
FIGURE D3.1.2.12 Calcification of the papillary muscles in an ex vivo CT scan. Massive calcification on the anterior papillary muscle (a) and less distinctive on the posterior one (b) at the formalin-fixed autopsy specimen. By CT there are one dense spot of calcification (c) and small hyperdensities (d) seen within the left ventricle.
Incident-Specific Cases
237
A
B
C
FIGURE D3.1.2.13 Endocarditis with vegetation on the aortic valve. A T2-weighted long-axis view of the left-ventricular outflow tract (A) demonstrates the large vegetation at the posterior leaflet (arrow). View of the aortic valve at autopsy from above (B) and after opening of the left-ventricular outflow tract (C) clearly depicts the dimensions of the vegetation.
enhancing agent application using the blood flow for its distribution to visualize ischemic myocardium. Thereby, the characteristics of enhancement are used to assess the size and age of the ischemia [34]. Postmortem cardiac imaging does not have the possibility to assess late enhancements. However, it provides an the excellent image quality due to the absence of any cardiac motion or breathing-related artifacts. Nevertheless, already by the early 1980s unenhanced clinical MRI also showed promising results for the identification of myocardial infarction using the significant increase of water content (edema) within the infarcted muscle [35,36]. These early studies as well as ones that followed showed that there is a significant prolongation of the T2 relaxation time within the infarcted myocardium [37–40]. T1 relaxation time was likewise prolonged, but to a lesser extent [39,40]. Therefore the image contrast was greatest on T2-weighted images [39]. This is so far in absolute concordance with the first results from postmortem studies, which also stated that the contrast in myocardial infarction images was greatest in T2-weighted images [11]. Whereas in recent years clinical cardiac MRI has
© 2009 by Taylor & Francis Group, LLC
implemented different contrast agents, application schedules, and contrast-agent- and cardiac-motion-adapted sequences to assess cardiac pathology [3,41–43], postmortem cardiac MRI still depends on the imaging of the unenhanced structural alterations that occurred during the different stages of myocardial infarction. The first studies that concentrated on the postmortem cross-sectional findings of the heart showed exemplarily that myocardial infarction is also detectable in postmortem MRI [11,44]. There are several publications that investigated the morphological alterations during and after myocardial ischemia and their appearances in postmortem MRI [9,45,46]. As cardiac deaths represent the major portion of natural deaths in the First World, it is of particular importance that postmortem imaging can detect especially these pathologies of the human heart. Macromorphological alterations such as hypertrophy or dilatation are less challenging in this context (see Chapter D3.1.3) than the tissue alterations occurring during and after cardiac ischemia [11]. However, as these are the most prevalent causes for cardiac deaths regardless of insufficiency or arrhythmia, these tissue alterations are needed to undergo further postmortem imaging research. As shown in previous studies [11,28], CT is an excellent tool for the detection of coronary calcifications also postmortem but fails in assessing critical tissue alterations of the myocardium. However, MRI can close this gap [10,11,44]. Regarding the forensic importance of this issue, additional questions need to be answered. Is early (peracute) myocardial infarction detection possible using postmortem MRI, which has been a challenge in traditional autopsy for decades [47–49]? Are there sequences that allow an improved assessment of myocardial infarction in postmortem imaging compared with the so far used T2-weighted sequences? To address these questions, different stages of myocardial infarction with different PMIs were investigated and correlated to the histological alterations within the myocardium. This was performed under consideration of the particular case histories in a well-directed study [45]. The aim was to create a time course of the postmortem imaging appearances of myocardial infarction similar to histological time courses that had already been compiled several decades prior [50,51]. First, immunohistochemical signs of myocardial infarction become visible after 2–4 hours survival time. Ischemia results in an early global acidosis (pH less than 6.9) due to the consumption of the ATP [52]. Thereby, the ischemic myocytes become increasingly susceptible to eosinophilic stains as a first visible sign in routine histology. In this time frame the occurrence of contraction band necrosis can be observed [53]. Single fibers or small groups of fibers show already a loss of striation and are swollen (cellular edema) or thinned and wavy [50,54]. Usually there is a small subendocardial layer (0.3–0.5 mm) of surviving tissue that is explained by an assumed oxygen diffusion through the endocardium [51,54]. Thereby, in two situations such as intraventricular thrombi covering the endocardium and
238
The Virtopsy Approach
pathologically thickened endocardium, no surviving subendocardial layer can be expected [55]. As a frequent finding, hemorrhages occur focally within the infarcted tissue. The inflammatory reaction is commenced by the infiltration of polymorphonuclear leucocytes. The amount of infiltration increases progressively up to the fourth day. After the fifth day the polymorphonuclear leucocytes become necrotic and begin to disappear. Simultaneously, mononuclear cells start the removal of the necrotic muscle by phagocytosis. Newly formed blood capillaries can be found that start growing into the infarcted regions from the periphery to the ischemic
center. Along these ingrowing vessels, fibroblasts reach the necrotic areas. The fibroblasts (fibroblast-like cells or myofibroblasts as they express alpha-smooth muscle actin) start the formation of collagen (type I and III) [51,56–58]. The formation of collagen is moderately prominent at three weeks and reaches its maximum at about three months. This fibrous tissue reaction also encroaches noninfarcted myocardial regions, but to a distinctively lesser extent [59–61]. All these ongoing alterations seem to be influenced only by an early revascularization during the first 4–6 hours after the coronary occlusion, resulting in a reduction of the infarct size but
A
B
FIGURE D3.1.2.14 The cardiac section technique was adapted to short-axis images (A) to increase the comparability of signal alteration and myocardial finding. This included histological specimen from the entire circumference of the left-ventricular myocardium (B).
© 2009 by Taylor & Francis Group, LLC
Incident-Specific Cases
239
not in a complete reversibility [62,63]. Later reperfusion does not influence the ongoing necrosis but accelerates subsequent inflammatory and remodeling reactions. It was expected that the well-defined histological alterations cause specific signal changing in postmortem MRI. The investigation of these signal alterations would allow for an infarction-age estimation based on the postmortem MR images. In 8 cases in the Virtopsy Project that presented with myocardial infarction as cause of death, 11 separately infarcted areas were investigated in situ using a 1.5 Tesla system. In addition to the normal axial, sagittal, and coronar thoracic scanning planes, the hearts of the investigated bodies were scanned using short, vertical, and horizontal long-axis images [12]; T1-weighted (echo time [TE] 15 ms; repetition time [TR] 400 ms) 2 mm slice thickness; T2-weighted (TE 96 ms; TR 4 s) 2 mm slice thickness (with and without fat saturation); short tau inversion recovery (STIR) (TE 14 ms; inversion time [TI] 133 ms; TR 3 s) 3 mm slice thickness and fluid-attenuated inversion recovery (FLAIR) (TE 217.5 ms; TI 2200 ms; TR 11002 ms) 3 mm slice thickness. Histological correlation was performed after histological specimens of the entire circumference of the left ventricle according to the short-axis MR images were prepared (Figure D3.1.2.14). Histological stains included hematoxylin and eosin (H&E), van Gieson, and chromotrop-aniline-blue (CAB). The coronary orifices and the apex were used as anatomical landmarks that allowed comparing similar slice sections in imaging and autopsy. Histological grading was performed according to
Table D3.1.2.1 from peracute, acute (early and late), subacute (early and late), and chronic. The appendant antemortem anamneses have been examined to assess subjective or objective signs for the time point of onset . This information combined with the histological age staging was used to determine the age of the ischemia. D3.1.2.4.4.1 Peracute Infarction Two cases of death due to a peracute myocardial ischemic event showed no myocardial morphological alterations at autopsy or in the applied histological stains. However, the diagnosis could be made as an indicative anamnesis was present and a fresh thrombotic coronary occlusion was found at autopsy. Thereby, both autopsy and imaging failed to visualize myocardial changes in peracute infarction. This is caused by the lacking survival time that would allow for vital reactions within the myocardium such as edema and necrosis. When ischemia and death of the individual occur in close succession, these reactions are not possible anymore. Even routine histology fails to discriminate between initial ischemic myocardium and subsequent death with then “global ischemic” myocardium. In these cases one still depends on coronary diagnostics showing stenosis or occlusions. Unfortunately, only half of these peracute cases present with an acute coronary lesion [64]. The remaining cases demonstrate extracardiac signs of cardiac failure but no pathologic cardiac findings. If the 3T scanners become more widespread, the higher resolution (matrix 1024 r 1024)
TABLE D3.1.2.1 Correlation of Myocardial Signal Alteration to the Histological Findings in Eight Investigated Myocardial Infarction Cases Case
1
Localization
inf
Peracute Acute
Early acute
Late acute Subacute
Early subacute Late subacute
Chronic
MRI signal
T1 T2 T2 fat saturation Stir Flair
Increased eosinophilia Wavy fibres Oedema Contraction bands Myocardial haemorrhage Neutrophiles infiltration Fibre necrosis Macrophages infiltration Fibroblasts Fibroblasts/fibrocytes Loose connective tissue Angiogenesis Fibrocytes Collageneous scar Hypertrophic adjacent fibres
2 lat
x
inf
lat
3
4
5
ant/sep
lat
inf
x
x
x x x x x x x
x x x x
ant
x x x x x x x
pp pp pp p
8
lat/inf
sep
x x x x x x x
0
7
x x x
x x x
cen 0 mar r cen rr mar pp cen rr mar pp cen rr mar pp cen r mar p
6 Global
cen 0 mar r cen rr mar pp cen rr mar pp cen rr mar pp cen r mar p
x x x
r
0
0
r
0
0
r
0
0
r
0
0
r
0
0
Basal/apical 0 Near rupture r cen rr mar pp cen rr mar pp cen rr mar pp cen r mar p
x x x x x
x x x
x x x x x x
r
0
0
rr
r
Varying
pp
rr
r
Varying
pp
rr
r
Varying
pp
rr
r
Varying
p
rr
x indicates the histological occurrence of the finding; cen, central; mar, marginal; 0, no signal alteration visible; p, slight increase in signal; pp, distinctive increase in signal; r, slight decrease in signal; rr, distinctive decrease in signal.
© 2009 by Taylor & Francis Group, LLC
240
The Virtopsy Approach
T1
T2
Stir
Flair
FIGURE D3.1.2.15 Acute myocardial infarction in postmortem MRI. Short-axis images near the cardiac apex show a very slight decrease but no further obvious signal alteration in the T1-weighted image. T2-weighted, STIR, and FLAIR (less distinctive) images present with a central signal decrease (yellow arrows) and increasing signal surrounding the hypointensity. This hyperintense margin is more noticeable on the epicardial margin (yellow dashed arrows). Histological appearance is dominated by a necrotic center of the lesion with eosinophilic fibers without nuclei and contraction band necrosis (left) surrounded by oedematous myocardium and infiltrating leucocytes (right).
might allow for the detection of smaller (e.g., edematous myocardial) alterations that can easily be overseen at autopsy but can nonetheless induce arrhythmias. Furthermore, recent research efforts concentrate on the possibilities of increasing the significance of postmortem imaging by an application of postmortem contrast agents [28,29]. As already shown in porcine ex vivo experiments, distribution defects of injected gadolinium could be simulated within the porcine myocardium [28]. This might allow for dealing with these forensic cases in a minimally invasive manner if necessary in the near future (Figure D3.1.2.10). Besides contrast-agent injection as a promising adjuvant technique supporting postmortem imaging, diffusion-weighted
© 2009 by Taylor & Francis Group, LLC
cardiac imaging might allow for imaging of fiber alterations and especially wavy fibers in early acute stages [65–68]. D3.1.2.4.4.2 Acute Infarction Four cases presented with acutely infarcted regions. Acute infarction (Figure D3.1.2.15, Figure D3.1.2.16, and Figure D3.1.2.17) with histological central necrosis, peripheral edema, and cellular reactions showed two different areas of signal behavior. A signal reduction within the necrotic center of the myocardial wall in T2-weighted, STIR, and FLAIR sequences was present. Predominantly subepicardial marginal regions showed increased signals in T2-weighted,
Incident-Specific Cases
241
T1
T2
Stir
Flair
FIGURE D3.1.2.16 In contrast to Figure D3.1.2.15, the T1-weighted image presents with a local signal increase within the infarcted inferior wall (arrows). Histology demonstrates broad intramural hemorrhage within the necrosis.
STIR, and FLAIR sequences. T1-weighted imaging failed to show distinctive signal alterations in the affected myocardial regions unless intramyocardial postinfarction hemorrhage was present. This caused the signal in T1-weighted images to increase slightly. The low signal of the necrotic center within the T2-weighted images might be a result of the hypoperfusion that led to fewer protons within these tissue regions (this is a hypothesis). Edema and cellular reactions starting from peripheral regions caused the signals in T2-weighted, STIR, and FLAIR images to be higher in marginal parts of the ischemic lesion. The proportion of necrotic-central signal decrease to reactivemarginal signal increase might be a useful hint for a more detailed estimation of the survival time in acute lesions. Of course, this needs to be investigated in a larger population of acute infarction cases. Postinfarction intramyocardial hemorrhage was the only histological finding that accompanied an
© 2009 by Taylor & Francis Group, LLC
increasing signal in T1-weighted images. Therefore, hyperintensities within T1-weighted images in regions that show acute infarction signs in T2-weighted images are strongly indicative for intramyocardial hemorrhage. D3.1.2.4.4.3 Subacute Infarction Four cases showed subacute infarcted regions. Subacute ischemia (Figure D3.1.2.18), histologically characterized by fibroblasts that form loose connective tissue replacing necrotic fibers and ingrowing vessels, showed increased signals in T2-weighted, STIR, and FLAIR images. T1-weighted images failed to show signal alterations in these regions. D3.1.2.4.4.4 Chronic Infarction Four cases showed chronically infarcted areas. Chronic infarction (Figure D3.1.2.19) with definite collagen formation
242
The Virtopsy Approach
T2
Stir
FIGURE D3.1.2.17 Postinfarction transmural myocardial rupture (yellow arrows) with pericardial tamponade (yellow dashed arrows) is obvious in all MR images. At autopsy there is a blood-filled pericardial space after opening of the pericardium (white arrow). Both, the formalin-fixed and the unfixed autopsy specimen present with a transmural rupture of the inferior wall (white dashed arrows).
T2
Stir
FIGURE D3.1.2.18 Subacute myocardial infarction in postmortem MRI. Short-axis images show a local hyperintensity in T2 and STIR within the posterior/lateral wall (yellow arrows) well correlating to the appearance of the local subacute subepicardial infarcted area at autopsy (yellow dashed arrow). Histology (H&E) demonstrates loose connective tissue formation and angiogenesis as the reason for the signal increase in T2-weighted sequences.
© 2009 by Taylor & Francis Group, LLC
Incident-Specific Cases
243
TABLE D3.1.2.2 Time Course of MR Signal in Postmortem MRI of Myocardial Infarction Hyperintense T1
Isointense Hypointense Necrotic centre Hyperintense
T2
Isointense Hypointense
Hyperintense T1
Isointense Hypointense Marginal regions Hyperintense
T2
Isointense Hypointense Peracute (0–~6h)
Acute (~1d–~1week)
Subacute (~1week–~2month)
Chronic (>~2month)
Note: T1 and T2 signal course within the necrotic center (upper graphs) and in marginal myocardial regions (lower graphs). Dashed line at the T1 graph of the necrotic center indicates the signal increase if intramural hemorrhage occurs. Signal course of STIR and FLAIR sequences behave similar to T2 but in distinctively lesser amplitudes.
at histology showed a signal loss in all applied sequences with distinctiveness decreasing from T2, STIR, FLAIR to T1. Subacute and chronic infarctions reach forensic importance as a cause of death due to a continuous reduction of the ejection fraction leading to acute cardiac insufficiency or by generating a lethal ventricular tachycardia. Around the electrically unexcitable collagen of myocardial scars, ventricular reentry tachycardia can occur [69–71]. Thereby, these findings are essential in about 40% of cardiac deaths when extracardiac findings such as congested internal organs indicate a cardiac failure as cause of death but acute ischemic myocardial findings are missing [64]. Based on the aforementioned study results, it was possible to create a time course of the postmortem MRI signal behavior correlated to the age of infarction at the time of death (Table D3.1.2.2). Looking at the combination of T1 and T2 signal within central and marginal parts of the myocardial lesions it became feasible to estimate the age of infarction. Obviously, the major limitation of the presented study was the low number of included myocardial infarction cases. As a forensic institution predominantly investigating homicides, suicides, accidents, and deaths that occurred under unclear circumstances, the very common cardiac death due
© 2009 by Taylor & Francis Group, LLC
to myocardial infarction is not that common in forensic case material. Close collaborations with pathological institutes or cardiologic clinics could result in larger study populations. Since the first CT and MR scanners were installed in forensic institutes in Copenhagen, Melbourne, and Bern, further studies are possible with less logistic efforts in the near future. To our knowledge, until now there have been no experiences with other myocardial alterations such as myocarditis in postmortem MRI. But these have to be taken into consideration as potential differential diagnoses and need to be addressed in specific future studies to define their postmortem imaging appearance. As long as the postmortem diagnosis in forensic medicine is based on a combination of imaging and image-guided biopsy with histological investigation, there is no risk for misinterpretation of images. D3.1.2.4.5 Hypertrophy and Dilatation Macromorphological alterations are precisely documented by postmortem MRI and even in CT [72]. Hypertrophy and dilatation can be measured on the short-axis view and additionally on the horizontal and vertical long-axis view by manually tracing the endocardial and epicardial contours at the
244
The Virtopsy Approach
T1
T2
Stir
Flair
A
B
FIGURE D3.1.2.19 Chronic myocardial infarction in postmortem MRI. Short-axis images show broad decrease in signal along the septum with thinning of the septal wall (yellow arrows). Autopsy demonstrates definite collagenous transformation of the infarcted septal myocardium (yellow dashes arrow) with scar-caused shrinking of the septal diameter. Histology at the border between scar and vital myocardium shows cell-free collagen (right) as the cause for the significant decrease of signal in MRI.
workstation (Figure D3.1.2.20). Similar to clinical radiology, it is possible to calculate the weight of normal-size as well as hypertrophied ventricles by multiplying the myocardial volume (epicardial minus endocardial volume, including papillary muscles and septum, excluding epicardial fat) with the factor 1.05 g/cm2 as the assumed density of the myocardium [73–75]. This offers the possibility of comparing obtained LV masses with data of normal LV masses acquired on MRI [76,77] and defining hypertrophy in consideration of gender differences. This can give even more precise and detailed information about LV hypertrophy than the often used entire heart weight measured at autopsy, which is compared with the so-called critical heart weight of 500 g. The latter always contains valve structures, parts of the great vessels, and above all the epicardial fat, which shows high interindividual differences and should not be included. Further morphologic relevant alterations such as local hypertrophies as seen in hypertrophic obstructive cardiomyopathia (HOCM) are easily visible in short-axis as well as in long-axis views (Figure D3.1.2.20) and can precisely
© 2009 by Taylor & Francis Group, LLC
FIGURE D3.1.2.20 Calculation of the LV muscle volume (a). On short-axis T2-weighted images, the left-ventricular areas are demarcated manually. The myocardial areas of all acquired slices are then multiplied by the gap between the slices and thereby integrated to estimate the entire left-ventricular volume. Further multiplying of volume by an assumed density of 1.05 g/ml leads to the LV mass. The result can be correlated to the normal range of LV mass by MRI. Hypertrophic obstructive cardiomyopathy (HOCM) (b). Short-axis view of T2-weighted MR sequence demonstrates an asymmetric hypertrophy of the anterior septum and the anterior wall of the left ventricle with a diameter of more than 25 mm. Note a hyperintense subacute septal myocardial infarction (arrow) as a typical finding in HOCM.
be measured. Myocardial subacute infarction is also seen in these regions as some local signal increase (Figure D3.1.2.20) [78–80]. Note an example of a massive biventricular hypertrophy as well—the so-called cor bovinum—in Figure D3.1.2.21. Also, right-ventricular alterations become imaged. Short-axis images can, for example, visualize right-ventricular dilatation (Figure D3.1.2.22). Detailed analysis of the cardiac MR images will reveal discrete wall alterations as seen in cases
Incident-Specific Cases
A
B
FIGURE D3.1.2.21 Cor bovinum in postmortem MRI. Dilatative cardiomyopathy of uraemic causation presents with massive dilatative hypertrophy of both ventricles. At autopsy a heart with a weight of more than 1 kg was found that confirmed the imaging diagnosis.
of right-ventricular dysplasia, such as fatty infiltration of the free right-ventricular wall and thinning of the apical rightventricular wall (Figure D3.1.2.23) [81–83]. D3.1.2.4.6 Injury Besides the head, the heart is an organ often aimed at for suicide as well as for homicide (e.g., knife wounds, gun shots). Therefore, imaging of the injuries of the heart is of high importance. Transmural injuries with hemopericardium are easily visualized by the macromorphological impression of an enlarged space between epicardium and pericardium filled with the same intensity of the signal as intravascular or intraatrial (Figure D3.1.2.24). The injury itself can be visualized depending on the size, location, and readaptation of the wound margins. Particularly in the absence of vital wound reactions
© 2009 by Taylor & Francis Group, LLC
245
A
B
FIGURE D3.1.2.22 Right-ventricular hypertrophy and dilatation in a case of a chronic pulmonary hypertension caused by a pulmonary granulomatosis. Note the thickening of the right-ventricular wall.
because of a sudden death and with closely readapted wound margins, visualization can become difficult. D3.1.2.5 Conclusion One major function of forensic medicine is to discriminate natural from nonnatural deaths. To establish a minimally invasive postmortem imaging procedure as an alternative to traditional autopsy, postmortem cross-sectional imaging must be able to serve as tool for assessing macromorphological changes and be able to replace the naked eye in a traditional autopsy. It is therefore thought to act similar to the eye of the forensic pathologist as a detection tool for local macropathological alterations and may allow an initial diagnosis but will normally depend on histological confirmation. Tissue specimen of pathological areas can be obtained using an image-guided biopsy technique [15], and today coronary diagnostics can be performed using minimally invasive postmortem angiography techniques [28,29]. As these angiography techniques are implemented based on MSCT, minimally
246
The Virtopsy Approach
A
invasive case management will require a combination of both cross-section techniques such as described in the virtopsy idea [10,29]. Especially peracute infarction cases depend on angiographic coronary diagnostics to visualize stenotic alterations or occlusions. D3.1.2.6 Acknowledgments The presented research results would not be possible without the motivated support of different persons. We would especially like to thank Elke Spielvogel MTRA (Med. Rad. Technician), Christoph Laeser MTRA, Carolina Dobrowolska MTRA, Karin Zwygart MTRA, and Verena Beutler MTRA for their excellent help during data acquisition, as well as Urs Königsdorfer and Roland Dorn for their experienced assistance with autopsy.
B
D3.1.2.7 References
FIGURE D3.1.2.23 Right-ventricular dysplasia in postmortem imaging. On T2-weighted MR images, fatty infiltration on a shortaxis image (arrow in A) and apical wall thinning (arrows in B) of the right ventricle on a long-axis image wall are demonstrated.
1. Simonetti, O. P., R. J. Kim, D. S. Fieno, et al. 2001. An improved MR imaging technique for the visualization of myocardial infarction. Radiology 218:215–23. 2. Simonetti, O. P., J. P. Finn, R. D. White, G. Laub, and D. A. Henry. 1996. “Black blood” T2-weighted inversion-recovery MR imaging of the heart. Radiology 199:49–57. 3. R. J. Kim, D. S. Fieno, T. B. Parrish, et al. 1999. Relationship of MRI delayed contrast enhancement to irreversible injury, infarct age, and contractile function. Circulation 100:1992–2002. 4. Becker, A., A. Leber, C. W. White, C. Becker, M. F. Reiser, and A. Knez. 2007. Multislice computed tomography for determination of coronary artery disease in a symptomatic patient population. Int J Cardiovasc Imaging 23:361–67. 5. Scheffel, H., H. Alkadhi, A. Plass, et al. 2006. Accuracy of dual-source CT coronary angiography: First experience in a high pre-test probability population without heart rate control. Eur Radiol 16:2739–47.
FIGURE D3.1.2.24 Knife wound through the apical part of the left ventricle. A short-axis T2-weighted image of the cardiac apex shows the myocardial defect as the cardiac part of the stab wound (large arrow) and the voluminous effusion causing cardiac tamponade (small arrows). At autopsy the transmural stab wound is confirmed.
© 2009 by Taylor & Francis Group, LLC
Incident-Specific Cases
6. Savino, G., P. Zwerner, C. Herzog, et al. 2007. CT of cardiac function. J Thorac Imaging 22:86–100. 7. Ko, S. M., J. B. Seo, M. K. Hong, et al. 2006. Myocardial enhancement pattern in patients with acute myocardial infarction on two-phase contrast-enhanced ECG-gated multidetector-row computed tomography. Clin Radiol 61:417–22. 8. Jackowski, C. 2003. Macroscopical and histological findings in comparison with CT- and MRI- examinations of isolated autopsy-hearts. Thesis, Institute of Forensic Medicine, O.-v.-G.-University of Magdeburg. 9. Schweitzer, W., K. Yen, M. Thali, et al. 2001. Documentation of heart-findings of 35 autopsies in comparison to CT and MRI. Paper (V-82) presented at the 80th annual meeting of DRGM, Interlaken. 10. Thali, M. J., K. Yen, W. Schweitzer, et al. 2003. Virtopsy, a new imaging horizon in forensic pathology: virtual autopsy by postmortem multislice computed tomography (MSCT) and magnetic resonance imaging (MRI)—a feasibility study. J Forensic Sci 48:386–403. 11. Jackowski, C., W. Schweitzer, M. Thali, et al. 2005. Virtopsy: postmortem imaging of the human heart in situ using MSCT and MRI. Forensic Sci Int 149:11–23. 12. Axel, L. 1992. Efficient method for selecting cardiac magnetic resonance image locations. Invest Radiol 27:91–93. 13. Jackowski, C., M. Thali, E. Aghayev, et al. 2006. Postmortem imaging of blood and its characteristics using MSCT and MRI. Int J Legal Med 120:233–40. 14. Imaizumi, T., M. Chiba, T. Honma, and J. Niwa. 2003. Detection of hemosiderin deposition by T2*-weighted MRI after subarachnoid hemorrhage. Stroke 34:1693–98. 15. Aghayev, E., M. J. Thali, M. Sonnenschein, C. Jackowski, R. Dirnhofer, and P. Vock. 2007. Post-mortem tissue sampling using computed tomography guidance. Forensic Sci Int 166:199–203. 16. Adams, V. and C. Guidi. 2001. Venous air embolism in homicidal blunt impact head trauma. Case reports. Am J Forensic Med Pathol 22:322–26. 17. Messmer, J. M. 1984. Massive head trauma as a cause of intravascular air. J Forensic Sci 29:418–24. 18. Kerner, T., G. Fritz, A. Unterberg, and K. J. Falke. 2003. Pulmonary air embolism in severe head injury. Resuscitation 56:111–15. 19. Adams, V. I. and C. S. Hirsch. 1989. Venous air embolism from head and neck wounds. Arch Pathol Lab Med 113:498–502. 20. Mallach, H. J. 1978. Air embolism as a primary or a concurrent cause of death. Hefte Unfallheilkd 132:52–55. 21. Mallach, H. J. and W. K. Schmidt. 1980. The quantitative and qualitative procedure for the determination of gas or air embolisms. Beitr Gerichtl Med 38:409–19. 22. Erben, J. and F. Nadvornik. 1963. The quantitative demonstration of air embolism in certain cases of fatal trauma. J Forensic Med 10:45–50. 23. Stanford, W. 1999. Coronary artery calcification as an indicator of preclinical coronary artery disease. Radiographics 19:1409–19. 24. Kajinami, K., H. Seki, N. Takekoshi, and H. Mabuchi. 1995. Noninvasive prediction of coronary atherosclerosis by quantification of coronary artery calcification using electron beam computed tomography: comparison with electrocardiographic and thallium exercise stress test results. J Am Coll Cardiol 26:1209–21.
© 2009 by Taylor & Francis Group, LLC
247
25. Keelan, P. C., L. F. Bielak, K. Ashai, et al. 2001. Long-term prognostic value of coronary calcification detected by electron-beam computed tomography in patients undergoing coronary angiography. Circulation 104:412–17. 26. Margolis, J. R., J. T. Chen, Y. Kong, R. H. Peter, V. S. Behar, and J. A. Kisslo. 1980. The diagnostic and prognostic significance of coronary artery calcification. A report of 800 cases. Radiology 137:609–16. 27. Rah, B. R., R. J. Katz, A. G. Wasserman, and J. S. Reiner. 2001. Post-mortem three-dimensional reconstruction of the entire coronary arterial circulation using electron-beam computed tomography. Circulation 104:3168. 28. Jackowski, C., M. Sonnenschein, M. J. Thali, et al. 2005. Virtopsy: postmortem minimally invasive angiography using cross section techniques--implementation and preliminary results. J Forensic Sci 50:1175–86. 29. Jackowski, C., S. Bolliger, E. Aghayev, et al. 2006. Reduction of postmortem angiography-induced tissue edema by using polyethylene glycol as a contrast agent dissolver. J Forensic Sci 51:1134–37. 30. Kizer, J. R., W. B. Gefter, A. S. deLemos, B. J. Scoll, M. L. Wolfe, and E. R. Mohler, III. 2001. Electron beam computed tomography for the quantification of aortic valvular calcification. J Heart Valve Dis 10:361–66. 31. Gottdiener, J. S. and W. C. Roberts. 1998. Severe mitral regurgitation late after healing of myocardial infarction from calcification of the posteromedial left ventricular papillary muscle. Am J Cardiol 81:662. 32. Madu, E. C. and I. A. D’Cruz. 1997. The vital role of papillary muscles in mitral and ventricular function: echocardiographic insights. Clin Cardiol 20:93–98. 33. Schwender, F. T. 2001. Papillary muscle calcification after inferoposterior myocardial infarction. Heart 86:E8. 34. Ibrahim, T., S. G. Nekolla, M. Hornke, et al. 2005. Quantitative measurement of infarct size by contrast-enhanced magnetic resonance imaging early after acute myocardial infarction: comparison with single-photon emission tomography using Tc99m-sestamibi. J Am Coll Cardiol 45:544–52. 35. Herfkens, R. J., R. Sievers, L. Kaufman, et al. 1983. Nuclear magnetic resonance imaging of the infarcted muscle: a rat model. Radiology 147:761–64. 36. Higgins, C. B., R. Herfkens, M. J. Lipton, et al. 1983. Nuclear magnetic resonance imaging of acute myocardial infarction in dogs: alterations in magnetic relaxation times. Am J Cardiol 52:184–88. 37. Garcia-Dorado, D., J. Oliveras, J. Gili, et al. 1993. Analysis of myocardial oedema by magnetic resonance imaging early after coronary artery occlusion with or without reperfusion. Cardiovasc Res 27:1462–69. 38. Krauss, X. H., E. E. van der Wall, A. van der Laarse, et al. 1990. Follow-up of regional myocardial T2 relaxation times in patients with myocardial infarction evaluated with magnetic resonance imaging. Eur J Radiol 11:110–19. 39. McNamara, M. T., D. Tscholakoff, D. Revel, et al. 1986. Differentiation of reversible and irreversible myocardial injury by MR imaging with and without gadolinium-DTPA. Radiology 158:765–69. 40. Karolle, B. L., R. E. Carlson, A. M. Aisen, and A. J. Buda. 1991. Transmural distribution of myocardial edema by NMR relaxometry following myocardial ischemia and reperfusion. Am Heart J 122(3 Pt 1):655–64.
248
41. Krombach, G. A., M. F. Wendland, C. B. Higgins, and M. Saeed. 2002. MR imaging of spatial extent of microvascular injury in reperfused ischemically injured rat myocardium: value of blood pool ultrasmall superparamagnetic particles of iron oxide. Radiology 225:479–86. 42. Steuer, J., T. Bjerner, O. Duvernoy, et al. 2004. Visualisation and quantification of peri-operative myocardial infarction after coronary artery bypass surgery with contrast-enhanced magnetic resonance imaging. Eur Heart J 25: 1293–99. 43. Fieno, D. S., R. J. Kim, E. L. Chen, J. W. Lomasney, F. J. Klocke, and R. M. Judd. 2000. Contrast-enhanced magnetic resonance imaging of myocardium at risk: distinction between reversible and irreversible injury throughout infarct healing. J Am Coll Cardiol 36:1985–91. 44. Jachau, K., T. Heinrichs, W. Kuchheuser, et al. 2004. Computed tomography and magnetic resonance imaging compared to pathoanatomic findings in isolated human autopsy hearts. Rechtsmedizin 14:109–16. 45. Jackowski, C., A. Christe, M. Sonnenschein, E. Aghayev, and M. J. Thali. 2006. Postmortem unenhanced magnetic resonance imaging of myocardial infarction in correlation to histological infarction age characterization. Eur Heart J 27:2459–67. 46. Shiotani, S., K. Yamazaki, K. Kikuchi, et al. 2005. Postmortem magnetic resonance imaging (PMMRI) demonstration of reversible injury phase myocardium in a case of sudden death from acute coronary plaque change. Radiat Med 23:563–65. 47. Varga, M. and L. Zsonda. 1988. A simple method for postmortem detection of acute myocardial infarction. Forensic Sci Int 37:259–63. 48. Hansen, S. H. and K. Rossen. 1999. Evaluation of cardiac troponin I immunoreaction in autopsy hearts: a possible marker of early myocardial infarction. Forensic Sci Int 99:189–96. 49. Ribeiro-Silva, A., C. C. S. Martin, and M. A. Rossi. 2002 Is immunohistochemistry a useful tool in the postmortem recognition of myocardial hypoxia in human tissue with no morphological evidence of necrosis? Am J Forensic Med Pathol 23:72–77. 50. Fishbein, M. C., D. Maclean, and P. R. Maroko. 1978. The histopathologic evolution of myocardial infarction. Chest 73:843–49. 51. Mallory, G. K., P. D. White, and J. Salcedo-Salgar. 1939. The speed of healing of myocardial infarction—a study of the pathologic anatomy in seventy-two cases. Am Heart J 18:647–71. 52. Walpoth, B. H., J. Galdikas, A. Tschopp, et al. 1991. Differentiation of cardiac ischemia and rejection by nuclear magnetic spectroscopy. Thorac Cardiovasc Surg 39 Suppl 3:217–20. 53. Baroldi, G., R. E. Mittleman, M. Parolini, M. D. Silver, and V. Fineschi. 2001. Myocardial contraction bands. Definition, quantification and significance in forensic pathology. Int J Legal Med 115:142–51. 54. Fishbein, M. C., D. Maclean, and P. R. Maroko. 1978. Experimental myocardial infarction in the rat: qualitative and quantitative changes during pathologic evolution. Am J Pathol 90:57–70. 55. Bouchardy, B. and G. Majno. 1974. Histopathology of early myocardial infarcts. A new approach. Am J Pathol 74:301–30. 56. Sun, Y. and K. T. Weber. 2000. Infarct scar: a dynamic tissue. Cardiovasc Res 46:250–56. 57. Morales, C., G. E. Gonzalez, M. Rodriguez, C. A. Bertolasi, and R. J. Gelpi. 2002. Histopathologic time course of myocardial infarct in rabbit hearts. Cardiovasc Pathol 11:339–45.
© 2009 by Taylor & Francis Group, LLC
The Virtopsy Approach
58. Virag, J. I. and C. E. Murry. 2003. Myofibroblast and endothelial cell proliferation during murine myocardial infarct repair. Am J Pathol 163:2433–40. 59. Volders, P. G., I. E. Willems, J. P. Cleutjens, J. W. Arends, M. G. Havenith, and M. J. Daemen. 1993. Interstitial collagen is increased in the non-infarcted human myocardium after myocardial infarction. J Mol Cell Cardiol 25:1317–23. 60. Litwin, S. E., C. M. Litwin, T. E. Raya, A. L. Warner, and S. Goldman. 1991. Contractility and stiffness of noninfarcted myocardium after coronary ligation in rats. Effects of chronic angiotensin converting enzyme inhibition. Circulation 83:1028–37. 61. Morales, C., M. Rodriguez, G. E. Gonzalez, M. Matoso, C. A. Bertolasi, and R. J. Gelpi. 2001. Time course of the myocardial infarction in the rabbit. Medicina (B Aires) 61:830–36. 62. Smith, G. T., J. R. Soeter, H. H. Haston, and J. J. McNamara. 1974. Coronary reperfusion in primates. Serial electrocardiographic and histologic assessment. J Clin Invest 54:1420–27. 63. Geft, I. L., M. C. Fishbein, J. Hashida, et al. 1984. Effects of late coronary artery reperfusion after myocardial necrosis is complete. Am Heart J 107:623–29. 64. Farb, A., A. L. Tang, A. P. Burke, L. Sessums, Y. Liang, and R. Virmani. 1995. Sudden coronary death. Frequency of active coronary lesions, inactive coronary lesions, and myocardial infarction. Circulation 92:1701–09. 65. Scollan, D. F., A. Holmes, R. Winslow, and J. Forder. 1998. Histological validation of myocardial microstructure obtained from diffusion tensor magnetic resonance imaging. Am J Physiol 275(6 Pt 2):H2308–18. 66. Geerts, L., P. Bovendeerd, K. Nicolay, and T. Arts. 2002. Characterization of the normal cardiac myofiber field in goat measured with MR-diffusion tensor imaging. Am J Physiol Heart Circ Physiol 283:H139–45. 67. Hsu, E. W., A. L. Muzikant, S. A. Matulevicius, R. C. Penland, and C. S. Henriquez. 1998. Magnetic resonance myocardial fiber-orientation mapping with direct histological correlation. Am J Physiol 274(5 Pt 2):H1627–34. 68. Chen, J., S. K. Song, W. Liu, et al. 2003. Remodeling of cardiac fiber structure after infarction in rats quantified with diffusion tensor MRI. Am J Physiol Heart Circ Physiol 285(3):H946–54. 69. Soejima, K., W. G. Stevenson, W. H. Maisel, J. L. Sapp, and L. M. Epstein. 2002. Electrically unexcitable scar mapping based on pacing threshold for identification of the reentry circuit isthmus: feasibility for guiding ventricular tachycardia ablation. Circulation 106:1678–83. 70. Stevenson, W. G. and E. Delacretaz. 2000. Strategies for catheter ablation of scar-related ventricular tachycardia. Curr Cardiol Rep 2:537–44. 71. Vester, E. G. 1998. Myocardial ischemia and ventricular arrhythmia. Z Kardiol 87(Suppl 2):49–60. 72. Shiotani, S., M. Kohno, N. Ohashi, et al. 2003. Dilatation of the heart on postmortem computed tomography (PMCT): comparison with live CT. Radiat Med 21:29–35. 73. Marcus, J. T., L. K. DeWaal, M. J. Gotte, R. J. van der Geest, R. M. Heethaar, and A. C. van Rossum. 1999. MRIderived left ventricular function parameters and mass in healthy young adults: relation with gender and body size. Int J Card Imaging 15:411–19. 74. Rajappan, K., L. Livieratos, P. G. Camici, and D. J. Pennell. 2002 Measurement of ventricular volumes and function: a comparison of gated PET and cardiovascular magnetic resonance. J Nucl Med 43:806–10.
Incident-Specific Cases
75. Jauhiainen, T., V. M. Jarvinen, and P. E. Hekali. 2002. Evaluation of methods for MR imaging of human right ventricular heart volumes and mass. Acta Radiol 43:587–92. 76. Lorenz, C. H., E. S. Walker, V. L. Morgan, S. S. Klein, and T. P. Graham, Jr. 1999. Normal human right and left ventricular mass, systolic function, and gender differences by cine magnetic resonance imaging. J Cardiovasc Magn Reson 1:7–21. 77. Alfakih, K., S. Plein, H. Thiele, T. Jones, J. P. Ridgway, and M. U. Sivananthan. 2003. Normal human left and right ventricular dimensions for MRI as assessed by turbo gradient echo and steady-state free precession imaging sequences. J Magn Reson Imaging 17:323–29. 78. Bogaert, J., M. Goldstein, F. Tannouri, J. Golzarian, and S. Dymarkowski. 2003. Original report. Late myocardial enhancement in hypertrophic cardiomyopathy with contrast-enhanced MR imaging. AJR 180:981–85. 79. Kim, R. J. and R. M. Judd. 2003. Gadolinium-enhanced magnetic resonance imaging in hypertrophic cardiomyopathy: in vivo imaging of the pathologic substrate for premature cardiac death? J Am Coll Cardiol 41:1568–72. 80. Wilson, J. M., R. P. Villareal, R. Hariharan, A. Massumi, R. Muthupillai, and S. D. Flamm. 2002. Magnetic resonance imaging of myocardial fibrosis in hypertrophic cardiomyopathy. Tex Heart Inst J 29:176–80. 81. Di Cesare, E. 2003. MRI assessment of right ventricular dysplasia. Eur Radiol 13(6):1387–93. 82. Tandri, H., H. Calkins, K. Nasir, et al. 2003. Magnetic resonance imaging findings in patients meeting task force criteria for arrhythmogenic right ventricular dysplasia. J Cardiovasc Electrophysiol 14:476–82. 83. Harper, K. W. and R. Tello. 2003. Prediction rule for diagnosis of arrhythmogenic right ventricular dysplasia based on wall thickness measured on MR imaging. Comput Med Imaging Graph 27:363–71.
D3.1.3 OTHER ORGAN-SYSTEM PATHOLOGIES Lars Oesterhelweg and Michael J. Thali D3.1.3.1 Vascular System In cases of natural death due to vascular pathologies, three major entities are of relevance: ischemia, thrombosis, and hemorrhages. The major entities of death by ischemia are related to the heart and the brain and are described in previous chapters. D3.1.3.1.1 Thrombembolism The primary locations of fatal thrombembolisms are the pulmonary arteries. Pulmonary embolism is a commonly seen cause of death in forensic medicine, and it is a life-threatening disease, which appears in up to 20% of all hospitalized patient [1–5]. Many conditions may lead to a thrombosis, and it is still a field of ongoing research; however, the thrombogenic triad of Rudolf Virchow (*1821 †1902), which is more than 100 years old, is still reliable. Virchow gave three factors for the appearance of thrombosis: (1) injuries of the vessel wall; (2) hypocirculation (stasis); and (3) hypercoagulopathy of the blood. These conditions are present in nearly all surgical hospitalized patients. Stasis is also present in individuals who sit in a narrow space for a long period of time, such as on a lengthy flight—which leads to the vivid term
© 2009 by Taylor & Francis Group, LLC
249
economy class syndrome [6,7]. Individuals who suffer from a pulmonary embolism may present symptoms like tachycardia, chest pain, and syncopy or respiratory distress. These symptoms are not specific and are also present in many other diseases. About one third of all major pulmonary embolisms are fatal within one hour. Other risk factors or predispositions are female gender, oral contraceptives or pregnancy, old age, obesity, malignancies, venous insufficiencies, and hereditary coagulopathies (e.g., factor-V-Leiden, protein-C- and protein S-deficiencies, AT-III deficiency) [8]. The most common origin (90%) of thrombemboli-causing pulmonary embolism is the deep veins of the lower extremities. As clinical signs of a deep-vein thrombosis of the lower extremity are regularly reported but are only present in the minority of patients, a difference in the circumference of the legs is one of the signs accessible in the external postmortem investigation. In many cases the mechanism of death seems to be natural, but due to the cause of events the classification is accident, suicide, or even homicide. If a patient dies from pulmonary embolism it could be due to hereditary diseases or confinement to bed due to previous assault or accident. Death from pulmonary embolism is caused by an obstruction of the pulmonary arteries, which leads to a reduction of the blood flow and consecutive right-heart failure. On the other hand, this obstruction leads to a further vasoconstriction with a loss of the diameter of the pulmonary arteries. In clinical radiology, the diagnosis of pulmonary embolism is performed by contrast-enhanced spiral computed tomography (CT) with a sensitivity of 60% and a specificity of 95% [8] (Figure D3.1.3.1). In conventional radiology, a large majority of patients have abnormalities in the chest x-ray, such as atelectasis, pleural effusion, pleural opacity, elevated diaphragm, decreased vascularity, and cardiomegaly. Postmortem radiological diagnosis remains difficult today. These difficulties are based on the postmortem imaging of blood [10]. The presence of postmortem blood clots in the major vessels and the development of sedimentation of the corpuscular blood components and of internal livores in the lung tissue are the major difficulties. A secure differentiation between postmortem blood clots and real thrombemboli is often difficult by autopsy alone, and histological procedures are necessary for a valuable diagnosis. In postmortem radiology the signs of right-heart failure and pleural effusion are detectable by multislice CT (MSCT) and magnetic resonance imaging (MRI), but today the differentiation between thrombemoli and postmortem coagula or sedimentation remains difficult [10]. There is still a great need for research in this field of postmortem imaging. Postmortem angiography and image-guided biopsy seem to be possibilities to diagnose pulmonary embolism in the deceased, but scientific evaluation remains necessary. D3.1.3.1.2 Hemorrhages Hemorrhages by natural causes appear by rupture of arteries due to aneurysms. While intracranial aneurysms are described in Chapter D3.6, “Fatal Hemorrhage in Postmortem Cross-Sectional
250
The Virtopsy Approach
aneurysms of the aorta with a five-year rupture risk of 50% [12]. In younger patients, aneurysms are often related to collagenous diseases such as Marfan syndrome or Ehlers-Danlos syndrome or cystic medial necrosis. Also syphilis and cocaine consumption should be kept in mind in the differential diagnosis [13,14,15]. In clinical and postmortem MSCT and MRI, hemorrhages caused by rupture of major vessel aneurysms are easily visible, as demonstrated in Chapter D3.6. Also, the diameter of the major arteries is easily measurable, and medium-sized and large aneurysms (Figure D3.1.3.1) as well as hemorrhages to the retroperitoneal soft tissue are visualizable. D3.1.3.2 Respiratory System (a)
Despite the cerebral, cardiac, or vascular causes of death, fatal respiratory failures are mostly caused by chronic or progressive diseases and are rare in forensic medicine. Sudden death due to diseases of the respiratory tract is most often caused by asthma or pneumonia [8]. D3.1.3.2.1 Asthma
(b) FIGURE D3.1.3.1 Clinical, contrast-enhanced CT scan with an incidental finding of an aortic aneurysm in a 71-year-old patient: (a) Axial cross-section with a maximum diameter of 5 cm. (b) Sagittal cross-section with a maximum length of 6 cm.
Radiology,” aortic aneurysms appear with a sudden cause of death due to intrathoracic or intra-abdominal hemorrhages. In two thirds of cases, the etiology of aneurysm is based on arteriosclerosis. Other etiologies are congenital, inflammatory, and traumatic [11]. In clinical radiology, aneurysms of the aorta are a frequent coincidental finding in routine ultrasound, x-ray, CT, or MRI examinations. The predominant location of these aortic aneurysms is abdominal and most often infrarenal. In 5% of all men between the ages of 65 and 79, aneurysms could be detected [12]. Abdominal aneurysms are mostly of atherosclerotic origin. When an aortic aneurysm ruptures, there is a more than 50% mortality rate. In literature, a significant rise in spontaneous ruptures is noted at a diameter of 6 cm for abdominal
© 2009 by Taylor & Francis Group, LLC
Asthma affects about 3% of the Western population and might also be a cause of unexpected death [16]. Sudden death from asthma is rare and may occur with a death rate from 1.1% to 7%. Frequency of death from asthma is increased at night or in the early morning and must not be affiliated with a prolonged attack. In 25% of death from asthma, the patient dies within 30 minutes after the onset of the attack [2]. In an asthma attack, the acute mechanism of death is acute rightheart failure, but also unintentional fatal intoxication by an overdose of inhalative sprays should be mentioned [17]. Due to the air trapping in the lungs, they usually do look overexpanded at time of autopsy and are occupying the complete chest cavity. This autopsy finding of acute emphysema may not be present after cardiopulmonary resuscitation. In addition, white, sticky, tenacious mucus is present in the bronchi [2,8]. So in postmortem imaging the diagnosis of a sudden death due to asthma is not possible today. In these cases only the sign of acute emphysema—if there was no extended resuscitation—might be visible in cross-section imaging. D3.1.3.2.2 Pneumonia In classififying pneumonia, different schemes are used. There are topographic (interstitial, lobular, lobar), etiologic (bacterial, viral, fungal, parasitic, toxic), and chronologic (acute, chronic) classifications used in the clinical routine (Table D3.1.3.1). Infections of the bronchial tract are common diseases, and pneumonia is the leading cause of death in hospitalized patients. In these patients it must be considered that a reason of hospitalization might be of nonnatural origin, such as a traffic accident. On the other hand, there are certain condition like intoxication and other consciousness conditions that can easily lead to an aspiration of stomach contents and subsequent pneumonia. The classification of a natural or nonnatural cause of death is difficult in all of these cases [2,8,18].
Incident-Specific Cases
251
Pneumonia as an unexpected cause of death without hospitalization occurs often in individuals with an impaired immune system, which is often present in the homeless, in alcoholism, acquired immune deficiency (e.g., HIV infection, cancer, therapeutically immune suppression), or old age. In these cases lobar pneumonia is common. The lethality of lobar pneumonia is more than 50% in patients over the age of 50 [8]. In clinical radiology, chest x-ray is still the method of choice in the diagnosis of pneumonia because infiltrations with a higher density are visualizable. In postmortem radiology the infiltrations are also detectable by postmortem CT and MRI. In Figure D3.1.3.2(a–c), postmortem CT and MRI
TABLE D3.1.3.1 Classification of Pneumonia Topography
Etiology
Chronology
Lobular Lobar Interstitial
Bacterial Viral Fungal Parasitic Toxic
Acute Chronic
(a)
(b)
(c) FIGURE D3.1.3.2 Images of the lungs of a 94-year-old man, who died 6 days after a traffic accident (pedestrian versus car) due to increasing intracranial pressure: (a) Axial CT scan of the thorax with infiltrations in the left lungs. (b) Coronar MRI scan (loc) with infiltration in the left lower lobe of the lungs. (c) Autopsy findings in the lower lobe of the left lung including pus in some bronchi and some scattered foci of consolidation.
© 2009 by Taylor & Francis Group, LLC
252
The Virtopsy Approach
(a)
(c)
(b) FIGURE D3.1.3.3 A 52-year-old woman who was found dead in her apartment. She had a history of alcohol abuse and known obesity with the therapy of gastric banding: (a) Axial abdominal postmortem MSCT scan, which demonstrates the signs of liver cirrhosis with hepatomegaly, inhomogenicity, and nodular margins of the liver (red arrow) as well as signs of significant blood loss with a collapsed aorta (yellow arrow). Also visible in this image is the gastric banding (green arrow). (b) Autopsy finding of liver cirrhosis. (c) Autopsy finding of a lesion in an esophageal varix near the cardia.
© 2009 by Taylor & Francis Group, LLC
Incident-Specific Cases
as well as autopsy images of a lobular pneumonia in the lower part of the left lung are shown. Sometimes it is difficult to differentiate between the pneumonial infiltration and internal livores. In corpses that are found in a position other than supine and are stored or transported in the supine position, the remains of internal livores may be visible on radiological imaging as an inhomogenicity in the lungs. D3.1.3.3 Gastrointestinal Tract Gastrointestinal causes of sudden death are mostly of hemorrhagic nature. Gastric or duodenal ulcers as well as esophageal varices are the leading entities. The diagnosis of gastrointestinal bleeding is not that difficult because hematemesis and melena are easily detectable by an external examination, but forensic interest can be focused on the origin of bleeding as well as by large amounts of blood on the corpse and its surroundings. Especially in massive bleedings from the upper gastrointestinal tract fresh red blood may be present at the scene at large. Regarding the manner of death, it must be considered that previous operations in the mouth and pharynx, like tonsillectomy or tooth extraction, may cause massive bleeding with a delay of up to a few days. Also, gastroscopies may cause injuries and can be the cause of a massive bleeding [5]. In these cases a nonnatural death must be certified. A major cause of death from the upper gastrointestinal tract in forensic medicine is bleeding from esophageal varices, which occur in individuals with portal hypertension. In most of these individuals cirrhosis of the liver is present due to chronic alcohol abuse [18,20]. During autopsy, the spot-like origin of the bleeding is often difficult to identify because, due to the blood loss, the varices are collapsed. But in absence of other sources like ulcers of the upper gastrointestinal tract or gastric erosion, bleeding from esophageal varices is most likely in the presence of liver cirrhosis. Like at autopsy in postmortem radiology, the point of bleeding from the esophageal varices is not detectable today, but the signs of hemorrhage and cirrhosis of the liver can be detected (Figure D3.1.3.3[a–c]). In the future, postmortem angiography might be a valuable tool for the detection of these lesions. Other causes of gastrointestinal hemorrhages are peptic ulcers and cancer. In the Western world, extended diseases of undiagnosed cancer are uncommon, so these cases are rarely seen in forensic medicine. Peptic ulcers are often associated with Helicobacter pylori or nonsteroid antiinflammatory agents and analgesics [8,20]. These ulcers can vary in their diameter—usually less than 2 cm but occasionally much larger [20]. The most common complication of peptic ulcer is bleeding. This may be occult or, like in most of these cases in forensic medicine, a massive hemorrhage due to erosion of a gastric vessel. In clinical radiology the diagnosis of gastrointestinal ulcers remains difficult without oral-contrast agents, which are not available for postmortem radiology due to the loss of the peristaltic activity and the presence of gastrointestinal content in most corpses. But like in other causes of death with a massive blood loss, in postmortem radiology the signs of a massive hemorrhage as
© 2009 by Taylor & Francis Group, LLC
253
described in Section D3.6 are present, and in the future postmortem angiography might be able to detect the sources of bleeding in these cases. D3.1.3.4 References 1. Berzlanovich, A., M. Muhm, and G. Bauer. 2000. Cases of fulminating pulmonary embolism among autopsies performed at the Vienna Institute of Forensic Medicine (1983– 1992). Arch Intern Med 160:1530–35. 2. DiMaio, V. J. and D. DiMaio. 2001. Forensic Pathology, 2d ed. Boca Raton, FL: CRC Press. 3. Kürkciyan, I., G. Meron, F. Sterz, et al. 2000. Pulmonary embolism as a cause of cardiac arrest: presentation and outcome. Arch Intern Med 160:1529–35. 4. Lignitz, E., G. Lignitz, and K. Püschel. 1995. Todesursache Lungenembolie in der Rechtsmedizin. Versicherungsmed 47:203–07. 5. Saukko, P. and B. Knight. 2004. Knight’s Forensic Pathology. London: Arnold. 6. Cruickshank, J. M., R. Gorlin, and B. Jennet. 2007. Air travel and thrombotic episodes: the economy class syndrome. Lancet 2(8609):497–98. 7. Feltracco, P., S. Barbieri, F. Bertamini, E. Michieletto, and C. Ori. 2007. Economy class syndrome: still a recurrent complication of long journeys. Eur J Emerg Med 14:100–03. 8. Püschel, K. 2004. Plötzlicher Tod im Erwachsenenalter. In Handbuch gerichtliche Medizin, Ed. B. Brinkmann and B. Madea, 965–1060. Berlin: Springer Verlag. 9. Galvin, J.R. 2003. The diagnosis of pulmonary embolism. In Radiologic Pathology, vol. 1, 2d ed., ed. K. K. Koeller, A. D. Levy, P. J. Woodward, et al., 99–108. Washington, DC: American Registry of Pathology. 10. Jackowski, C., M. Thali, E. Aghayev, et al. 2006. Postmortem imaging of blood and its characteristics using MSCT and MRI. Int J Legal Med 120:233–40. 11. Riede, U. N., C. Ihling, and H. E. Schäfer. 1995. Arterien. In Allgemeine und spezielle Pathologie, 4. Auflage, Ed. U. N. Riede and H. E. Schäfer, 436–60. Stuttgart: Georg Thieme Verlag. 12. Dalsing, M. C., A. P. Sawchuk, S. G. Lalka, and E. R. Mohler. 1997. Vascular medicine. In Textbook of Internal Medicine, 3d ed., ed. W. N. Kelley, 473–88. Philadelphia: Lippincott-Raven. 13. Fikar, C. R. and S. Koch. 2000. Etiologic factors of acute aortic dissection in children and young adults. Clin Ped 39:71–80. 14. Rashid, J., M. J. Eisenberg, and E. L. Topol. 1996. Cocaine induced aortic dissection. Am Heart J 132:1301–04. 15. Davies, M. J., T. Treasure, and P. D. Richardson. 1996. The pathogenesis of spontaneous arterial dissections. Heart 75:434–35. 16. Benatar, S. R. Fatal asthma. 1986. New Engl J Med 314:423–39. 17. Morlid, I. and J. C. Giertsen. 1989. Sudden death from asthma. Forensic Sci Int 42:145–50. 18. Haleem, M. A. 1990. Aspiration pneumonia as a cause of death. Brit J Clin Pharmaco 44:398–99. 19. Riede, U. N., H. Denk, M. Stolte, and H. E. Schäfer. 1995. Hepatopankreatisches System. In Allgemeine und spezielle Pathologie, 4. Auflage, Ed. U. N. Riede and H. E: Schäfer, 742–96. Stuttgart: Georg Thieme Verlag. 20. Rubin, E. and J. L. Farber. 1994. The gastrointestinal tract. In Pathology, 2d ed., Ed. E. Rubin and E. L. Farber, 618– 702. Philadelphia: J.B. Lippincott.
254
The Virtopsy Approach
D3.2 POSTMORTEM IMAGING OF BLUNT TRAUMA Stephan A. Bolliger, Steffen Ross, and Michael J. Thali
D3.2.1 INTRODUCTION Injuries due to blunt trauma are a common finding in everyday forensic practice. The field of blunt trauma encompasses a huge variety of injuries, reaching from barely visible, superficial excoriations to the destruction of a corpse due to an impact of tremendous energy, such as in airplane accidents or persons run over by a train. Indeed, blunt trauma can be regarded as a cornerstone in forensic pathology. Often, internal injuries that are not evident during external inspection, no matter how thoroughly the examination was performed, may give rise to a certain moment of surprise at autopsy. The extent of externally visible damage does not always correlate to the extent of internal injuries. The realization of this disparity regarding external and internal findings made autopsies necessary and therefore led to the profession of the forensic pathologist. Although certain specially trained physicians, the forefathers of modern forensic pathologists, were performing autopsies for judicial systems several centuries ago—in 1532 A.D. under the Constitutio Criminalis Carolina of emperor Charles V of Germany—the introduction of medical x-ray examinations only took place at the beginning of the 20th century [1]. In the first radiographs performed for clinical reasons, a central question was whether suspected fractures could be verified and, if so, which type the confirmed fractures were. As forensic pathology addresses similar questions, the implementation of radiology in forensic pathology was a logical step in the assessment of blunt trauma. Today, hardly any institute for forensic pathology can exist without at least an x-ray machine for the assessment of fractures and foreign bodies. With modern machines and improved techniques and experience, not only osseous injuries but also lacerations of internal organs can be visualized. In the forensic assessment of victims of blunt trauma, the following points besides the possibility of the involvement of a third party are of utmost importance: r Type and extent of sustained injuries r Injury mechanism and identification of the injurycausing instrument r Cause of death r Reconstruction of the course of events r Vitality of the sustained injuries—that is, whether the injuries were inflicted prior to or after death As mentioned already, the field of blunt trauma is vast. Therefore, only a sample of the possibilities of postmortem imaging of such findings can be discussed and shown in this chapter. Postmortem imaging of vital reactions, which
© 2009 by Taylor & Francis Group, LLC
indicate that a person was alive upon infliction of the injuries, is discussed in Chapter D2.3, “Vital Reactions and Vital Signs,” and cerebrocranial trauma is discussed in Chapter D3.3.
D3.2.2 INJURY TYPE, MECHANISM, AND CAUSE OF DEATH D3.2.2.1
Blunt Trauma of the Superficial Skin Layers
Due to its most external location and extent, the skin is the organ that generally shows most signs of blunt trauma. These lesions range from superficial excoriations, intracutaneous hemorrhages, and subcutaneous hematomas to crush wounds and avulsions. These soft-tissue injuries are often easily detected at external inspection of the corpse. Depending on the wound morphology, a differentiation between blunt and sharp trauma can generally be made easily. Sometimes, an instrument or a weapon gives rise to distinct excoriations or intracutaneous hemorrhages by which the inflicting instrument can be identified. Such so-called patterned injuries are discussed in Chapter B1, “External Body Documentation.” As the external examination of these often extremely important superficial injuries is the method of choice and imaging does not give additional information to a thoroughly performed external inspection, this chapter does not deal with these lesions. D3.2.2.2 Blunt Trauma of the Subcutaneous Fatty Tissue and Musculature By contrast to what was said previously about superficial lesions of the skin, subcutaneous hematomas deserve further mentioning. These are not always evident at external inspection. Every physician has encountered patients or victims in which immediately after injury infliction no hematomas are seen. When patients or victims present themselves for a follow-up examination one or two days later, they then often display clear and sometimes even striking hematomas (Figure D3.2.1). This is due to the fact that hematomas need a certain amount of time to “bloom.” The speed of this hematoma blooming depends on a multitude of factors. Coagulopathy, skin thickness, and location on the body are just a few of these influencing factors. Several methods have been applied to visualize hematomas more adequately: At autopsy, suspected skin regions are incised. Obviously, this method is not applicable to living patients. Other methods are diaphanoscopy and spectrophotometry [2,3]. Having said this, it is obvious that corpses may not display external hematomas even after a severe blunt trauma (Figure D3.2.2 and Figure D3.2.3). This is especially true in cases of a rapid death. In previous studies [4], magnetic resonance imaging (MRI) proved to be a sufficiently sensitive method to detect softissue injuries. As subcutaneous hematomas and trauma of the fatty tissue are forensically relevant as to energy and impactpoint reconstruction, Yen et al. [5] evaluated such traumatic effects and graded them into four categories, namely, I–IV. All these stages or categories were clearly seen in MRI. The
Incident-Specific Cases
255
FIGURE D3.2.1 Photograph of the left thigh of a homicide suspect. She claimed to have stabbed her boyfriend in self-defense while he was beating and kicking her. The top image was taken a few hours after the incident, and the bottom image was taken two days later. Note the barely visible hematoma (encircled) on the photo taken on the day of the incident and the large, brown-yellow hematoma two days later.
mildest stage is I, in which only a perilobular hemorrhage is seen. Stage II implies a contusion of the fat and stage III a disintegration of the fat lobules. The most severe form of fatty-tissue injury is stage IV. In this stage, which is the result of an enormous local force against a body part, not only are fat lobules disintegrated, but also a subcutaneous cavity is encountered. Usually, the entire body has to be skinned at autopsy to assess the existence and the extent of such typical signs of blunt trauma. As Yen et al. have shown, postmortem MRI is a viable tool for such an assessment (Figure D3.2.4 and Figure D3.2.5). This technique is especially valuable in examining surviving victims of blunt trauma, where a pathological analysis is not possible. The next deeper structure of the body, namely the musculature, may also be injured in blunt injuries. Their injury implies that a greater energy was involved than necessary to merely crush the fat lobules of the subcutaneous fatty tissue. Hemorrhages into the muscles are hard to detect but are
© 2009 by Taylor & Francis Group, LLC
readily visible in MRI (Figure D3.2.6). Furthermore, a crushing of the muscle, an indicator that a considerable local force was applied, can also be visualized in MRI. D3.2.2.3
Blunt Trauma to the Head
Blunt injuries to the head can be differentiated into two groups: (1) impact injuries and (2) injuries due to a rapid change in velocity (acceleration and deceleration). Of the first group, soft-tissue injuries of the scalp and facial musculature (lacerations, abrasions, and contusions), skull fractures, cerebral contusions/lacerations, intracerebral hemorrhages, and epidural hemorrhages are notable. The second group, namely the change in velocity, which is, for instance, encountered in shaken baby syndrome and vehicle accidents, typically presents with subdural hematomas and diffuse axonal injuries. The latter is discussed in Chapter D3.3.4 on craniocerebral trauma. Although the topic of cerebral trauma is discussed in
256
The Virtopsy Approach
FIGURE D3.2.2 Autopsy photograph of the right hip side of a pedestrian who was hit by a car. Note the externally barely visible hematoma on the top (circle) and the crushing of the underlying subcutaneous fatty tissue on the lower image.
detail in Chapter D3.3, the importance of skeletal lesions to the head in blunt trauma deserves brief mentioning. D3.2.2.3.1
Direct Trauma of the Head
D3.2.2.3.1.1 Facial Fractures Facial fractures are a common finding in clinical forensic medicine and in forensic pathology. They usually arise due to direct trauma as in vehicle accidents, falls onto the face (often seen in intoxicated persons), or, more frequently, in fights and scuffles. Of these, the fracture of the nasal skeleton is the most frequently encountered. Although generally a self-limiting lesion with little or no danger to the individuals’ life or general health, nasal fractures can lead to the suspicion of an involvement of a third party in otherwise inconspicuous conditions. If an otherwise unharmed body of a young man is found in a locked flat with signs of a nasal fracture, then further investigations must be undertaken, even if the presence of a perpetrator at the time of death can be excluded. As the nasal fracture is a possible sign of a prior fight, an autopsy must be performed to examine the possibility of further, externally unseen lesions. However, a nasal fracture may be missed at external inspection. As a dissection of the face leads to disfigurement, the pathologist often refrains from this procedure. Therefore, the nasal fracture may even be missed after an otherwise complete autopsy has been performed.
© 2009 by Taylor & Francis Group, LLC
Postmortem multislice computed tomography (MSCT) easily detects such possibly telltale fractures (Figure D3.2.7). Other frequently seen fractures of the facial bones concern the eye sockets and can arise due to direct blunt trauma due to punches or impact from a flying object such as a ball. These
FIGURE D3.2.3 Autopsy photograph of the abdomen of a homicide victim who was kicked and throttled. The skin shows no injuries. However, upon incision, a crushing of the subcutaneous fatty tissue became evident.
Incident-Specific Cases
FIGURE D3.2.4 Postmortem imaging of a pedestrian who was hit by a car. The subcutaneous hematomas that are clearly depicted as signal-intense (white) regions in the lower image, namely a T2-weighted MRI, are not visible in the corresponding MSCT (upper image).
blow-out fractures are typically located at the medial and basal wall, where the bone is extremely thin [6,7]. However, eye socket fractures can also arise due to indirect trauma, such as in cases of falls with an impact to the back of the head as a contrecoup lesion [8]. Whereas these indirect fractures, typically located at the roof of the eye socket, are formed by a negative pressure in the fossa anterior, the opposite may also give rise to fractures of the orbita ceiling [9]. Such a rapid positive pressure can be achieved by gunshots to the skull. As is well known from clinical medicine, mid-face fractures can also extend over the maxilla and the zygomatic bone (Figure D3.2.8). These fractures arise from a direct impact to the face. Depending on the involved structures, these fractures are classified as Le Fort I–III. This classification can be difficult, especially in cases of vehicle accidents, where, due to the massive damage, multiple fractures are often seen. In such cases, the Le Fort classes overlap (Figure D3.2.9). Mandibular fractures occur due to punches, falls, and vehicle accidents, to name just a few mechanisms. Direct fractures are seen paramedially, whereas indirect fractures are mostly located in the region of the joint and the mandibular body.
© 2009 by Taylor & Francis Group, LLC
257
FIGURE D3.2.5 Axial T2-weighted MRI (top) and autopsy photo (bottom) of the right leg of a man who had been run and rolled over by a car. Note the signal-rich accumulation (arrow) of blood corresponding to the hemorrhage into the décollement lesion.
FIGURE D3.2.6 Postmortem MRI T2-weighted coronar image of the buttocks of a bicycle rider who was hit by a car. Note the feathery-appearing signal-intense (white) structures, which correspond to an intramuscular hematoma of the buttocks on the right.
258
The Virtopsy Approach
FIGURE D3.2.7 MSCT reconstruction of the skull of a woman who crashed her car against a tree. Note the barely visible nasal fracture (arrow) due to the dashboard impact against the face. The spiny formation in the mouth is due to dental artifacts.
Postmortem imaging can display such possibly telltale injuries in a rapid and nondestructive manner, thus sparing the face from further disfigurement or the pathologist from missing a potentially important finding. D3.2.2.3.1.2 Direct Trauma to the Skull Whereas in cases of sharp trauma the type of inflicting instrument may be discerned in most cases, this is not true for all contusions and lacerations of the scalp. For instance, it is obviously of utmost importance to distinguish between scalp contusions, due either to a fall or to a blow with an instrument. The fracture pattern of the skull and typical cerebral lesions can solve this problem with a large degree of certainty. If the head strikes a broad, flat surface, such as the ground, the skull is flattened at the point of impact. Due to this resulting inward bending, distant areas of the skull are bent outward. Fractures do not begin at the point of impact but at the point of outbending at the external surface [10,11]. For instance, a fall with a low-energy impact to the occipital skull will therefore classically lead to linear fractures (Figure D3.2.10). If the impact of the large, flat surface is great enough, complete or incomplete circular fractures may arise around the impact point at the edge of inward and outward bending. With an even greater amount of energy, the severe inbending at the point of impact leads to stellate fractures arising from the impact center. A combination of circular and stellate fracture lines creates a distinct spider web-like fracture system. Later fracture lines will not cross preexisting fracture lines, as the necessary tension is lacking in previously fractured areas. This phenomenon, also known
© 2009 by Taylor & Francis Group, LLC
FIGURE D3.2.8 MSCT 3D skull reconstruction of a young man who was struck by a falling tree. Besides the clearly visible nasal fracture (yellow arrow), several other fractures of the facial skeleton are seen (red arrows). The spiky structures in the region of the dentition are streak artifacts due to metallic components of dental restorations.
as the “Puppe-rule” can help assess the timing of skull injuries when more than one impact point is seen. If the impact occurs with high energy and a small surface area, i.e., in blows with hammers, the result is a small, depressed skull fracture (Figure D3.2.11 and Figure D3.2.12). Here, the brain is generally only affected in the immediate vicinity of the impact. A blow to the head (i.e., the occiput) will therefore mainly result in a cerebral injury to the occipital brain regions. Several blows to the head can obscure the small, depressed fracture due to the severe destruction of the concerned skull region. By contrast, a fall from great height or from an upright position onto the ground will lead to a completely different fracture pattern and cerebral injury (Figure D3.2.13). Here, linear or, if the energy involved is great enough, circular fractures are commonly encountered. Simple linear fractures are, however, also seen in (lowenergy) blows to the head. In these cases, the plain radiograph of the skull does not suffice to distinguish between a blow and a fall. Here, postmortem MSCT imaging can deliver quick and reliable results. By enabling the visualization of the brain, coup-contrecoup lesions can be detected. This constellation of impact-near and impact-far injuries is often seen in
Incident-Specific Cases
259
FIGURE D3.2.10 MSCT reconstruction of the skull of a man who slipped and fell backward onto the tarmac. Note the fracture of the occiput with radiating fracture lines to the skull base and the left temple (yellow arrows). The fractures lines appear smooth. This is due to the smoothing artifact of the computer-generated reconstruction.
FIGURE D3.2.9 Postmortem MSCT 3D bone reconstruction of a man who jumped off a 40-meter-high bridge and landed flat on his front. Note the extensive fracturing of the facial skeleton.
The relative weight of the head, together with the additional weight of the helmet, can tear the head from the neck in the case of a frontal collision. Torsion of the calvaria from the skull base (Figure D3.2.14 and Figure D3.2.15) has also been
falls. They are an absolute rarity in cases of homicidal blows to the head. Thus, by visualization of the brain, the pathologist may be able to discern between a blow and a fall even if the external wound morphology may be obliterated or hidden (e.g., due to secondary animal involvement such as ants). The cerebral injuries inflicted by blows and falls are discussed in detail in Chapter D3.3. D3.2.2.3.2 Indirect Blunt Trauma of the Head Indirect blunt traumas of the head encompass changes of velocity. The head is relatively heavy. The resulting inertia can lead to various lesions, typically of the skull base. These fractures of the skull base can be of great reconstructive value. If the fractured area is dislocated into the fossa posterior, the lesion must have occurred due to a deceleration of the head toward the spine in a stamp-like manner [12]. Such a constellation is seen in cases from falls from great heights such as suicidal falls onto the feet or in parachute accidents. The opposite is true for a rapid acceleration of the head from the spine. As the skull base is firmly attached to the spine by a multitude of ligaments, traction will result in a tearing of the skull base from the remaining cranium [13]. This rather rare finding is seen in cases such as motorcycle accidents.
© 2009 by Taylor & Francis Group, LLC
FIGURE D3.2.11 Autopsy photograph of a homicide victim who was clubbed to death with a baseball bat. Note the depressed region (yellow arrows) and the laceration of the right ear due to direct impact of the heavy instrument.
260
FIGURE D3.2.12 MSCT reconstruction of the skull of the case seen in Figure D3.2.11. Note the extensive destruction of the right side of the skull. The adjacent zygomatic bone is intact, thus underscoring the circumscript nature of the impact (in this case, blow) to the head.
FIGURE D3.2.13 MSCT reconstruction of the skull of a suicide victim who jumped off a 65-meter-high bridge. The left side of the skull shows a depressed fracture system. Fracture lines radiate to the right side of the skull. The spiny aspect of the molars is due to dental (metal) artifacts.
© 2009 by Taylor & Francis Group, LLC
The Virtopsy Approach
FIGURE D3.2.14 MSCT sagittal reformation of the head of a young man who suffered a tangential collision as a pedestrian with a train. The tangential collision resulted in a spinning away of the young man. This led to a rotation in the longitudinal axis as well as traction of the spine, which resulted in a ring fracture of the skull base. The arrows designate the fracture lines.
FIGURE D3.2.15 MSCT reconstruction of the skull. Here, the ring fracture of the skull base is depicted (arrows) after virtual removal of the skull cap.
Incident-Specific Cases
261
described as a mechanism for skull base ring fractures [14]. Another form of indirect blunt trauma of the head is seen in cases of shaken baby syndrome. This very controversial topic is discussed in Chapter D3.3. D3.2.2.4 Blunt Trauma to the Neck The neck is rightly deemed as being a particularly vulnerable region of the body. It is to be assumed that this is known to practically every perpetrator, regardless of his or her medical or anatomical knowledge. The vulnerability is due to the neck harboring several rather exposed, cerebrally vital blood vessels, the airways, and reflex centers. An excitation of the latter may lead to death from hypotension and cardiac arrest [15]. Blunt trauma to the neck consists of various, completely different mechanisms. High-velocity, direct impacts to the neck are seen in vehicle accidents and falls from great height. Here, the cervical lesions range from “mere” softtissue injuries to osseous lesions to complete decapitations. Of soft-tissue injuries, besides the easily externally detectable excoriations and sugillations, hemorrhages of the subcutaneous tissue and the cervical muscles are notable. The latter are extremely important in the forensic assessment of violent death due to blunt trauma. They are, however, not always visible at external examination, especially if the time between incident and death is very short. Indeed, an extremely important topic in the field of blunt trauma to the neck is choking, throttling, and hanging. Although abrasions and ligature marks are generally seen when an instrument such as a cord or rope is used (Figure D3.2.16 and Figure D3.2.17), such as in hanging and throttling, external signs of the neck may lack completely in cases of choking. In these cases, petechial hemorrhages in the face and mucosal linings of the head may give a first and vital clue as to the cause of death, thus (hopefully) leading to
FIGURE D3.2.16 Photograph of a homicide victim who was throttled to death with a belt. Note the accompanying abrasions (arrows) that occurred when the struggling victim tried to remove the belt.
© 2009 by Taylor & Francis Group, LLC
FIGURE D3.2.17 MSCT 3D surface rendering of a victim who committed suicide by hanging. The noose, or ligature, mark (arrows) is clearly visible.
a more thorough investigation. In cases of suicidal hanging, such petechial hemorrhages can be absent. Here, the position of the knot (i.e., behind the neck) and especially the course of the noose arms should be noted [16]. In cases of strangling or throttling, a limited amount of pressure is exerted onto the cervical vessels, thus giving rise to petechial hemorrhages. However, they are rarely encountered in cases of classic suicidal hanging. In these cases, the afferent and efferent blood vessels to the head are typically—if the noose is behind the mandibular angle—equally compressed. Therefore, no petechial hemorrhages are to be expected. Here, a thorough medicolegal inspection can reveal telltale imprints to the neck, even if the noose has been removed. Slight and superficial abrasions of the skin can often lead to the correct diagnosis. Additional hemorrhages of the subcutaneous tissue and the cervical musculature can, according to our experience, further support this hypothesis (Figure D3.2.18), although recent literature is uncertain as to the overall significance [17,18]. Furthermore, fractures of the hyoid bone or thyroid horns, a finding easily detected in postmortem MSCT (Figure D3.2.19 and Figure D3.2.20), give rise to the diagnosis blunt trauma to the neck [19,20]. These situations, when faced by inexperienced or all too experienced colleagues, can give rise to the premature diagnosis suicide by hanging. Unfortunately, it cannot be stressed
262
FIGURE D3.2.18 Coronal T2-weighted MR image of the neck of the victim seen in Figure D3.2.16. The signal-intense area at the base of the left side of the neck (arrow) is consistent with a hemorrhage.
FIGURE D3.2.19 MSCT 3D reconstruction of the thyroid cartilage and the hyoid bone. Note the fractures to these structures (arrows).
© 2009 by Taylor & Francis Group, LLC
The Virtopsy Approach
FIGURE D3.2.20 MSCT 3D bone reconstruction of the hanging victim seen in Figure D3.2.17. The left hyoid horn is fractured (arrow).
enough that scene findings, no matter how convenient they may be, should always be seen in a critical light in the context as a whole. If only a single finding does not fit into the whole scene, then the examination should be performed with even more ardor and scrutiny than usual. The question of a homicide by prior choking and subsequent hanging to imitate a suicidal hanging must always be addressed [21]. Apart from the crime scene investigation, the medicolegal examination is of utmost importance. Here, several techniques, such as the assessment of muscle histology [22], can be of assistance. As Aghayev et al. [23] show, the examination of the posterior cricoarytenoid muscle can serve as an imminently important finding in cases of suspected blunt trauma to the neck such as in choking. Postmortem imaging, especially in the form of MRI, can visualize such lesions that would otherwise only be detected at autopsy. The pathology and forensic imaging of choking is dealt with in further detail in Chapter D3.7, “Strangulation.” Obviously, osseous lesions are easily seen in postmortem MSCT. Hyoid and thyroid fractures can give clues as to whether a blunt trauma occurred to the neck. A caveat is nevertheless to be made: Anatomical variations of the thyroid and hyoid structures may give the impression of such violent injuries. In these cases, a surrounding hemorrhage—as found in traditional autopsies—must be looked for. The lack of such hemorrhages in postmortem imaging deems a vital or fresh fracture of the throat skeleton rather unlikely. Other lesions concern the cervical spine. Here, the vertebrae are of great interest. Almost every forensic pathologist has spent a considerable amount of time poring over
Incident-Specific Cases
FIGURE D3.2.21 MSCT sagittal reformation of a pedestrian who was hit by a car and delivered to the hospital in a tetraplegic state and died four days later. The dens axis is fractured (arrow) and slightly dislocated into the cervical canal.
2D, conventional x-rays of the neck. At external examination, the diagnosis of a cervical fracture can be quite tricky. The admitting diagnosis of probable hangman fracture on autopsy requests is often incorrect. This usually false diagnosis is probably due to the death-certifying physicians’ inexperience in examining corpses. In a fresh corpse, the lack of muscle tone compared with living persons may mimic a fracture of the cervical spine. A hangman fracture is—in contrast to the general layperson’s (and medical community’s) belief—a rarity. Such a fracture of the dens axis in a young and healthy individual with sufficient cervical muscle tone is indeed difficult to create. Such fractures arise only if a large amount of energy is applied to the neck in such a manner as to hyperextend the spine. This is usually only encountered in vehicle injuries (Figure D3.2.21, Figure D3.2.22, and Figure D3.2.23) and judicial hangings [24]. The latter should not be confused with the commonly seen suicidal hanging. In suicidal hangings, the victim usually sinks or falls (from a relatively low height) into the noose. In judicial hanging, seen in past days in Europe and still performed in certain countries (not just in the Third World), the victim falls from a relatively great height into the noose. Sometimes, weights are affixed to the victim in order to further increase the energy while falling. Often, the noose is tied below the chin, thus leading to an increased hyperextension of the neck. As opposed to judicial hanging, most forms of hanging do not suffice to break the neck. These almost always suicidal events can take place in various forms. Sometimes, the victims let themselves fall from a high structure such as a ladder into
© 2009 by Taylor & Francis Group, LLC
263
FIGURE D3.2.22 MSCT 3D bone reconstruction with virtual removal of the left portions of the skull of the case shown in Figure D3.2.21. The dens fracture (arrow) is clearly visible.
the noose. However, in certain cases the hanging can take place in a seated or even a lying position. The weight of the head alone suffices to generate enough traction on the rope to compress the (venous) backflow from the head, thus leading to a hypoxic brain death. Obviously, depending on the height of the fall into the noose—and therefore the energy involved—different injury
FIGURE D3.2.23 Sagittal T2 MR image of the head seen in Figure D3.2.21 and Figure D3.2.22. Note the discrete hypointensity of the spinal cord (arrows) at the level of the dens corresponding to a cervical lesion.
264
The Virtopsy Approach
patterns can arise. We have seen a case of complete decapitation when a man jumped off a 45-meter-high bridge with a long noose tied around his neck. In cases of seated or lying hangings, internal visual evidence for the hanging is often completely lacking. MSCT can easily reveal whether fractures of the spine, the hyoid, and the cartilaginous structures of the neck are present. On the other hand, MRI is better suited to visualize internal hemorrhages. D3.2.2.5 Thorax Due to the size of the thorax, blunt trauma often occurs to this body region. Blunt trauma can arise in the form of blows with an instrument, punches and kicks, an impact of a structure such as the ground in falls from great height, or the interior or exterior of a vehicle, to name just a few possibilities. Blunt trauma to the thorax often encompasses minor bruises and superficial excoriations, single or extensive rib fractures, and contusions and lacerations of the thoracic organs. Often encountered findings of little or no forensic relevance are postmortem rib fractures or fractures of the sternum. By external cardiopulmonary resuscitation, rib and (rarely) sternum fractures may arise. This is especially true in individuals with an underlying osteologic pathology such as osteoporosis. Not surprisingly, such fractures are commonly encountered in elderly patients. Usually, the upper half of the ventral portions of the ribs is involved. These are predominantly located on the left side of the chest, where the direct force in external cardiopulmonary resuscitation is applied. The hereby fractured ribs can occasionally puncture the pleura or the lungs. In a mistaken assumption of a cardiocirculatory arrest, this may lead to death due to pneumothorax or internal hemorrhage. However, these sometimes devastating complications are to be taken into account while resuscitating a lifeless person. The alternative of not trying to help—no matter how inexperienced or clumsy the resuscitating bypasser happens to be—is generally death. This view is obviously also shared by the judicial system: Refusal to help an injured person (with the exception of extremely rare circumstances) is punishable under the laws of most nations. Resuscitation-related injuries can be readily detected by the scarcity or lack of vital signs such as surrounding hemorrhages. MRI, the method of choice for soft-tissue injuries, can visualize such hemorrhages in clarity equal to traditional autopsy (Figure D3.2.24). Other important vital signs such as pulmonary fat embolism or, if the survival time is in excess of 30 minutes, inflammation can as yet not be sufficiently visualized in postmortem imaging (see Chapter D2.3, “Vital Reactions and Vital Signs”). A minimally invasive approach with biopsies of the organ systems most prone to such changes may serve as a viable compromise to the traditional autopsy. The effectiveness of sample gathering with MSCT-guided biopsy needles is being evaluated at present. Contusions of the subcutaneous fatty tissue and the underlying musculature are also frequently seen blunt injuries and have already been discussed.
© 2009 by Taylor & Francis Group, LLC
FIGURE D3.2.24 Pedestrian traffic accident victim. The upper image (3D MSCT bone reconstruction) shows multiple rib fractures (green arrows). These are surrounded by a considerable hemorrhage (red arrows, lower image, T2-weighted MRI)—indicating that they occurred premortem—as the signal-rich areas corresponding to intramuscular hemorrhages imply.
The next layer of the thorax is the rib cage. Here, the involved rib-fracturing mechanism is differentiated between indirect and direct force. In direct force to the thorax, the ribs in the immediate vicinity of the impact are broken. If the involved energy is great and localized, such as in blows with a heavy object or kicks, these ribs may reveal twofold or even more fractures (Figure D3.2.25). Such multiple fractures to a rib are a clear indication of a great amount of energy. Previously fractured ribs generally do not break further, as the necessary tension is destroyed by the preexisting lesion. By contrast, indirect fractures, as the name suggests, arise due to an impact at another site. By an extreme bending of the rib cage, the ribs break at the most curved—usually the
Incident-Specific Cases
FIGURE D3.2.25 MSCT reconstruction of the osseous structures of the thorax of a young woman who was kicked to death. Note the numerous, in some cases twofold, rib fractures (yellow arrows). These twofold fractures are typical for an impact of limited surface area and high energy, as would be the case in the (admitted) kicking of the victim.
lateral—portions. Such fractures are often seen in falls from great height due to the extreme deformation of the rib cage (Figure D3.2.26). They therefore serve to differentiate from the highly suspicious twofold (or even more) fractures. An exception to this general rule is, for example, the impact against an object, such as a tree branch, or if the victim’s own arm happens to be between the chest and the ground at the time of impact. This results in a transformation of a broad to a localized impact area and can thus lead to direct impact fractures (Figure D3.2.27). As mentioned already, regardless of the nature of the rib fractures, these can frequently lead to death. Death can arise due to internal hemorrhage, a laceration of the heart (Figure D3.2.28), the great blood vessels, or an intercostal artery, a pneumothorax due to puncturing of the lungs, or piercing of other internal organs by fracture splinters. A pneumothorax or even tension pneumothorax is obviously not visible at external inspection and can be difficult to detect at autopsy. There are several methods for the detection of such gaseous accumulations. First, the diaphragm is examined before dissection of the thorax. If it extends downward toward the abdominal cavity, a tension pneumothorax or a massive intrathoracic hemorrhage may be assumed. If the diaphragm location appears normal, then a tension pneumothorax—but not a “mere” pneumothorax—may be excluded. The latter can be visualized by puncturing the intercostal space and
© 2009 by Taylor & Francis Group, LLC
265
FIGURE D3.2.26 MSCT 3D reconstruction of a man who was crushed by a log. The multiple dorsally accentuated rib fractures are clearly visible and are located on the opposite side of the primary impact point, namely, the chest.
FIGURE D3.2.27 MSCT reconstruction of the left side of the rib cage of a victim of a fall from great height: Apart from a fracture of the left humerus, several left-sided rib fractures (arrows) are seen. The man hit the ground with the left side of the body. The left arm beside the chest upon impact transformed the otherwise broad impact region to a localized impact point, thus giving rise to this rib cage fracture distribution.
266
The Virtopsy Approach
FIGURE D3.2.29 MSCT coronal reconstruction of a traffic victim. The tension pneumothorax on the right (X) has displaced the mediastinal structures toward the left (arrows).
FIGURE D3.2.28 Axial images of a traffic accident victim. The MSCT shown on the top displays a large hemothorax (X); the corresponding T1-weighted MR is even clearer: Here, not only the hemothorax but also the blood sedimentation layers (yellow arrows) and the hemorrhage origin—namely, a cardiac rupture (green arrow)—are visible. The stripes in the MSCT are due to streak artifacts of glass and other foreign bodies within the body bag.
noting the position of the lungs. If the lungs sink into the thoracic cavity, then a relevant pneumothorax is hardly possible. For confirmation of a suspected tension pneumothorax, the following procedure is applied. By skinning the chest, a pocket between the outer chest wall and the skin is formed. If this pocket is filled with water and the chest is punctured, bubbles are created in the case of a tension pneumothorax. However, these methods are not always reliable, according to our experience. Furthermore, the extent of the pneumothorax cannot be evaluated. Postmortem MSCT solves this problem. Not only can a (tension) pneumothorax be proven or excluded beyond doubt (Figure D3.2.29), but also the amount of gas can be estimated. This is also true for tiny pneumothoraces. A special situation in blunt thoracic trauma is the formation of epipleural hematomas. By lesion of an intercostal artery and detachment of the parietal pleura from the chest wall, such an epipleural hematoma can form (Figure D3.2.30 and Figure D3.2.31). These hematomas can lead to death due to internal hemorrhage [25–27]. The radiological appearance on plain chest x-rays may be confusing and can lead to an
© 2009 by Taylor & Francis Group, LLC
incorrect diagnosis [28]. Furthermore, by opening the chest at autopsy, the membrane-like hematoma pocket may be damaged, and the correct diagnosis epipleural hematoma is thus made practically impossible. Postmortem MSCT, however, leads to an accurate visualization and thus recognition of the lesion.
FIGURE D3.2.30 MSCT coronar reformation of a traffic accident victim. The apical hyperdense cap of blood around the left pulmonary apex (X) corresponds to an extrapleural hematoma. Note also the pleural hemorrhage (arrows) in the left thoracic cavity.
Incident-Specific Cases
FIGURE D3.2.31 In this autopsy photograph of the case seen in Figure D3.2.30, the apex of the left thoracic cavity is shown. The parietal pleura is detached from the chest wall, thus giving rise to an extrapleural cavity (X).
If the exercising force is great enough, the thoracic deformation may lead to even greater lesions. The osseous structure of the rib cage of young adults, and especially children, leads to an enormous elasticity. Often, these individuals display no rib fractures but tremendous internal thoracic lesions. We have seen an infant who was killed by a reversing car, and the tire rolling over the thorax deformed it to such an extent that the lungs, the great blood vessels, and the heart were torn apart. Astonishingly, the infant lacked osseous lesions whatsoever. This elasticity of young and healthy individuals can be surpassed in cases where an enormous amount of energy is applied to the thorax. In cases of airplane crashes, accidents and suicides with trains, or high-velocity motor vehicle accidents and the rib cage can splinter. The speed and energy in these impacts do not allow for the rib cage to deflect the oncoming force sufficiently. Such a massive deformation can tear the heart, great blood vessels, and lungs apart. Additionally, the rib splinters can puncture these (and additional organs), resulting in multiple inner “stab” wounds. The problem in such cases is the rapid (almost immediate) death of the individual. Vital signs may lack completely at autopsy. Unfortunately, these are extremely important in the assessment of airplane or vehicle accidents, as the question of a possible prior (and therefore possibly accident-relevant) demise should always be addressed. In such cases, if no preexisting, potentially lethal pathologies are seen, one can cautiously postulate a trauma-related death. As every pathologist who has dealt with severely mutilated corpses has certainly experienced, this exclusion can be difficult or frustrating, as the evaluation of cardiac disease is hardly possible if the heart cannot be localized or is crushed beyond possible evaluation or even recognition.
© 2009 by Taylor & Francis Group, LLC
267
FIGURE D3.2.32 These two x-ray images of a body bag with the badly mangled remains of a suicide victim who was run over a by train do not sufficiently depict the contents.
Another problem is the evaluation of missing body parts. If a train runs and rolls over a person, the result may be severe carnage with body parts strewn over long distances. In one instance we have seen, the train dragged part of the scalp over 100 kilometers away. The severely destroyed body parts are often barely or not recognizable at all. However, it is of utmost importance to determine whether the main parts of the deceased have been found. If passersby find portions of, for instance, a hand or face a few days later, this will inevitably lead to the question of whether these parts did indeed belong to the previously found body—and therefore will lead to expensive DNA tests and of course to distress for whomever discovered the body parts. Although postmortem imaging is not able to reveal a great deal of additional information to the examination of the in certain cases almost dissected corpse, it is nevertheless suitable for a quick assessment of the available organs. This is especially true for badly mangled bodies delivered to the forensic department in body bags. By an MSCT scan, which according to our experience takes less than five minutes to perform, one can at least tell whether the main portions of the corpse are present and whether a further, more extensive, and time-consuming search must be undertaken (Figure D3.2.32, Figure D3.2.33, and Figure D3.2.34). D3.2.2.6 Abdomen and Pelvic Girdle D3.2.2.6.1
Abdomen
In blunt trauma, the abdomen has certain advantages and disadvantages compared with the thorax with regard to injury prevention. Although the abdomen lacks the protection of the rib cage, it generally does not suffer the stab wounds of the thorax when the rib cage splinters. Usually, the main lesions are tears or transections of the abdominal organs. These may lead to a
268
FIGURE D3.2.33 The MSCT surface reconstruction displays the remains of Figure D3.2.32 in a more suitable way. Portions of a rib cage, a hand, and a foot are clearly visible. However, these images do not show if or which body parts might still be missing and should therefore be sought after.
relevant or even fatal internal hemorrhage. Such organ lesions are often not visible externally. However, they and the resulting hemorrhages are easily detected in postmortem imaging (Figure D3.2.35 and Figure D3.2.36). Accumulations of free
FIGURE D3.2.34 Although this MSCT reconstruction lacks the three-dimensionality of Figure D3.2.32, it displays the body bag contents with regard to the skeletal remains in a suitable fashion. Here, large portions of the jaws, the right upper extremity, several parts of a rib cage, the severely destroyed pelvis, most of the right leg, and large portions of the left leg are clearly visible. The left arm is missing. This gave rise to a more thorough, eventually fruitful search.
© 2009 by Taylor & Francis Group, LLC
The Virtopsy Approach
FIGURE D3.2.35 Coronal T2-weighted MR image of the trunk. The signal-rich white region at the liver surface (yellow arrow) corresponds to a bleeding laceration of this organ, of which an autopsy photo is depicted as an insert. Note the signal-intense region at the left side of the abdomen (red arrow) due to a hematoma.
blood in the abdominal cavity are visible in both MSCT and MRI. As bleeding into the soft tissues is difficult to detect in MSCT, MRI is the visualization method of choice. As mentioned already, bleeding into a soft tissue (e.g., the fatty and muscle tissue) is seen as a signal-intense, and therefore brightwhite, region (Figure D3.2.37). The detection of fatal hemorrhages is discussed in Chapter D3.6, “Fatal Hemorrhage in Postmortem Cross-Sectional Radiology.”
FIGURE D3.2.36 MSCT coronal reformation of a victim who fell accidentally from scaffolding. Note the small laceration of the spleen (arrow), which corresponds to the autoptically proven splenic laceration (insert).
Incident-Specific Cases
FIGURE D3.2.37 Coronar T2-weighted MR image of a motorcycle driver who crashed into a bus. Note the extensive, bilateral signal-rich (white) regions (yellow arrows) corresponding to hemorrhages into the perirenal fat and the subcutaneous fatty tissue of the right arm (green arrow).
269
FIGURE D3.2.38 MSCT 3D bone reconstruction of a driver who sustained a frontal crash with his car. Note the shattered left femur and the fracture of the pubic bone (arrow).
D3.2.2.6.2 Pelvic Girdle The human pelvic girdle is special in the animal kingdom as it is a more or less rigid ring. A blunt force to one region of this ring will usually lead to a deformation of the pelvic girdle. This generally results in not only one but at least two fractures. Such fractures may arise due to direct impact as well as indirect impact. In direct impact, the force is, for example, a car bonnet in a vehicle accident or the ground in the case of a fall from great height with an impact against the pelvis. Indirect fractures are suffered when another body part transmits the sustained energy to the pelvic ring. Such a constellation is seen in cases of falls from great height onto the feet or in vehicle accidents involving the driver. In falls from great height the femur may be rammed through the pelvic structures into the abdominal cavity. In frontal vehicle accidents, the movement of the dashboard toward the knees of the driver results in a transmission of energy from the femur to the pelvis. Sometimes the femur breaks; in other cases the pelvis absorbs more energy and fractures (Figure D3.2.38). Fractures of the pelvic girdle may lead to fatal hemorrhages, to potentially lethal pulmonary fat embolism, or to organ damage due to bone splinters. D3.2.2.7 Extremities The extremities are among the most frequently damaged body parts. This applies not only to blunt trauma but also to other areas of forensic pathology, such as sharp trauma and gunshots. The extremities not only serve as a means of locomotion or as grasping and handling tools but can also be
© 2009 by Taylor & Francis Group, LLC
employed in aggressive or defensive actions. Therefore, injuries to the extremities—especially the upper extremities— should be examined closely, as these may indicate that the victim tried to ward off an attack and therefore would corroborate a perpetrator’s contention of self-defense. Furthermore, lesions of the extremities are very important in the assessment of the direction of the injury-inflicting force. The soft-tissue injuries previously discussed can indicate from which direction a person (i.e., a pedestrian) was struck by a vehicle, as crush pockets generally arise from a primary impact—that is, from the direct collision of the vehicle and not the subsequent fall. Osseous lesions such as a so-called Messerer wedge fracture of the long bones (Figure D3.2.39) are also a very important clue as to the side from which a pedestrian was struck [29]. Other telltale injuries of the extremities in blunt trauma concern the hands. Apart from the aforementioned defensive injuries to the hands, these parts of a body may also help to answer the question as to who steered an airplane at the time of the crash. Here, the hand of the pilot often grasps the steering clutch in a frantic attempt to avert the crash. Upon impact, the clutch is forced into the hand, thus giving rise to typical lesions (Figure D3.2.40, Figure D3.2.41, and Figure D3.2.42), which consist of tearing, often with accompanying hemorrhage, of the hand between thumb and index finger. Another less frequent site of hand injury is the region between the index and middle finger and the palm. Such telltale soft-tissue lesions can be detected using postmortem MRI, whereas MSCT is better suited for the detection of injuries to the skeletal structures. These revealing lesions
270
FIGURE D3.2.39 MSCT 3D reconstruction of the lower legs of a pedestrian who was hit by a car. The right fibula displays a wedgeshaped fracture (indicated by yellow lines).
FIGURE D3.2.40 Autopsy photograph of the right hand (palmar view) of an airplane pilot who crashed and burned. After removal of the charred skin, a hematoma of the palm (green arrow) is visible.
© 2009 by Taylor & Francis Group, LLC
The Virtopsy Approach
FIGURE D3.2.41 MSCT 3D bone reconstruction of the right hand of the pilot seen in Figure D3.2.40. Note the structure (arrow) grasped by the intact fingers, corresponding to a molten piece of the clutch.
FIGURE D3.2.42 T1-weighted image of the hand shown in Figure D3.2.40 and Figure D3.2.41. The hemorrhage seen in Figure D3.2.40 is clearly visible (green arrow), as are various other hemorrhages (red arrows).
Incident-Specific Cases
271
facilitating the presentation of findings to medical laypersons in court (Figure D3.2.43). Such whole-body presentations of MR scans are not possible with the current, state-of-the-art MR units due to the necessary long examination times. However, with new MR machines, whole-body MR scans have become possible (Figure D3.2.44.). The validation of one such system, the Total Imaging Matrix by Siemens, is currently under way.
D3.2.3 CONCLUSION
FIGURE D3.2.43 Whole-body MSCT reconstruction of the surface, the musculature, and the skeleton of a traffic victim.
can help the pathologist in the reconstruction of the course of the accident. In cases of falls from great height, fractures of the extremities in addition to the trunk and head injuries may provide useful information as to how the victim landed. Postmortem MSCT can give a good overview not only of the skeletal system but also of the skin and musculature of such victims in a nonbloody fashion. This is greatly helpful in
Postmortem imaging is useful in examining cases in which some form of blunt trauma was suffered, such as strangulation, a fall from great height, a traffic accident, or a homicidal clubbing. Postmortem MSCT is a valuable tool in detecting fractures and, in the case of whole-body scans, in demonstrating the fracture distribution in a three-dimensional fashion in a nonbloody form to medical laypersons and in facilitating their understanding of the case. Furthermore, small fractures in regions of the body that are difficult to access and that may be overlooked at autopsy can be detected easily. Another advantage is that certain areas of the body not routinely dissected, such as the face, can be examined with regard to fractures in a nondestructive fashion, thus sparing the next of kin more psychological trauma. On the other hand, MRI of victims of blunt trauma is well suited to detect lesions of the soft tissues such as the subcutaneous fatty tissue, the musculature, and the inner organs. By applying both methods, in combination with a 3D photogrammetry-based true-color surface scan for superficial injuries and biopsies for toxicological and histological analyses, a minimally invasive approach to the forensic examination of victims of blunt trauma is possible.
D3.2.4 REFERENCES
FIGURE D3.2.44 Whole-body MSCT reconstruction of the skeleton (left) and whole-body MRI of the same individual (middle and right image).
© 2009 by Taylor & Francis Group, LLC
1. Brogdon B.G. 1998. Forensic Radiology. Boca Raton, FL: CRC Press. 2. Horisberger B. and Krompecher T. 1997. Forensic diaphanoscopy: how to investigate invisible subcutaneous hematomas on living subjects. Int J Legal Med 110:73–78. 3. Bohnert M., Baumgartner R., and Pollak S. 2000. Spectrophotometric evaluation of the colour of intra- and subcutaneous bruises. Int J Legal Med 113:343–48. 4. Thali M.J., Yen K., Schweitzer W., Vock P., et al. 2003. Virtopsy, a new imaging horizon in forensic pathology: virtual autopsy by postmortem multislice computed tomography (MSCT) and magnetic resonance imaging (MRI)—a feasibility study. J Forensic Sci 48:386–403. 5. Yen K., Vock P., Tiefenthaler B., et al. 2004. Virtopsy: forensic traumatology of the subcutaneous fatty tissue; mutislice computed tomography (MSCT) and magnetic resonance imaging (MRI) as diagnostic tools. J Forensic Sci 49:799–806. 6. Goder G. and Simon G. 1985. Schädel-Hirn-Verletzungen aus der Sicht des Augenarztes. In: Lang G. and Reding R. (eds.), Schädel-Hirn- und Mehrfachverletzungen, 335–345. Leipzig: Barth.
272
7. Meier K. and Schmidt T. 1991. Diagnostische und therapeutische Aspekte bei medialen Orbitafrakturen. In: Walhart N.E. (ed.), Traumatologie des Mittelgesichts (Fortschritte der Kiefer- und Gesichtschirurgie, vol. 36), 220–22. Stuttgart: Thieme. 8. Prokop O. 1975. Einwirkung von stumpfer Gewalt. In: Prokop O. and Göhler W. (eds.), Forensische Medizin, 3d ed., 179–220. Berlin: Volk und Gesundheit. 9. Dettling J. 1938. Expressionsverletzungen des Schädels. Schweiz Med Wochenschr 68:486–89. 10. Gurdjian E.S., Webster J.E., and Lissner H.R. 1950. The mechanism of skull fractures. Radiology 54:313–38. 11. Lissner H.R. and Evans F.G. 1958. Engineering aspects of fractures. Clin Orthoped 8:310–22. 12. Bauer K. 1939. Der Bruch der Schädelbasis. Langenbecks Arch Klin Chir 196:460–514. 13. Voigt G.E. and Sköld G. 1974. Ring fractures of the base of the skull. J Trauma14:494–505. 14. Schulz E. and Jahn R. 1983. Ringfrakturen des Schädels. Z Rechtsmed 90:137–45. 15. Bolliger S., Plattner P., and Zollinger U. 2006. The magic broomstick: an unusual missile injury to the neck. Am J Forensic Path 27:304–06. 16. Plattner T., Yen K., Zollinger U., et al. 2004. Differenzierung von typischem und atypischem Erhängen. Wo liegt die praktische Bedeutung? Rechtsmedizin 14:266–70. 17. Kuznik J. and Keil W. 2000. The diagnostic significance of haemorrhage of the posterior cricoarytenoid miscles. Sci Justice 40:41–44. 18. Keil W., Kondo T., and Beer G.M. 1998. Haemorrhages in the posterior cricoarytenoid muscles: an unspecific autopsy finding. Forensic Sci Int 95:225–30. 19. Kokhlov V.D. 1996. The mechanisms of the formation of injuries to the hyoid bone and laryngeal and tracheal cartilages in compression of the neck. Sud Med Ekspert 39:13–16. 20. Betz P. and Eisenmenger W. 1996. Frequency of throatskeleton fractures in hanging. Am J Forensic Med Patho 17:191–93. 21. Mallach H.J. and Pollak S. 1998. Simulated suicide by hanging after homicidal strangulation. Arch Kriminol. 202:17–28. 22. Sigrist T., Germann U., and Markwalder C. 1997. Using muscle histology for assessment of vitality in hanging. Arch Kriminol 200:107–12. 23. Aghayev E., Jackowski C., Sonnenschein M., et al. 2006. Virtopsy: hemorrhage of the posterior cricoarytenoid muscle by blunt force to the neck in postmortem multislice computed tomography and magnetic resonance imaging. Am J Forensic Med Path 27:25–29. 24. Wallace S.K., Cohen W.A., Stern E.J., et al. 1994. Judicial hanging: postmortem radiographic, CT, and MR imaging features with autopsy confirmation. Radiology 193:263–67. 25. Bolliger S., Thali M.J., and Aghayev E. 2007. Post-mortem non-invasive virtual autopsy: extrapleural haemorrhage after blunt thoracic trauma. Am J Forensic Med Pathol 28:44–47. 26. Dirnhofer R., Sigrist T. and Ranner G. 1984. Epipleural haematoma. Etiology, morphology and clinical course. Unfallheilkunde 87:180–83. 27. Reiter C. and Denk W. 1985. Delayed epipleural hematoma as a fatal complication following blunt chest injury. Wien Klin Wochenschr 97:535–37.
© 2009 by Taylor & Francis Group, LLC
The Virtopsy Approach
28. Ranner G., Dirnhofer R., Sigrist T. et al. 1986. Unusual clinical forms of extrapleural (epipleural) hematoma on the chest x-ray. Radiologe 26:467–70. 29. Prokop O. and Dürwald W. 1975. Der tödliche Verkehrsunfall. In: Prokop O. and Göhler W. (eds.), Forensische Medizin, 208–20. Berlin: Volk und Gesundheit.
D3.3 FORENSIC NEUROIMAGING* K. Yen, J. Anon, L. Remonda, A. Spreng, and K. O. Lövblad
D3.3.1 INTRODUCTION The high relevance of neurotraumatology in forensic medicine is reflected by numerous statistics that rate intracranial injury as a main cause of death especially in younger people and as the main factor causing permanent disability [1–6]. Being today’s gold standard for the forensic postmortem examination of the head and brain, standard routine autopsy according to current guidelines is performed when it comes to evaluating intracranial findings [7]. In clinical radiology, the examination of head and brain injuries using cross-sectional imaging methods has a tradition of many years, and today detailed knowledge exists [8(and references therein),9]. This is not surprising, as most patients who suffer relevant minor or major head trauma undergo computed tomography (CT) and in some cases additional magnetic resonance imaging (MRI). Based on past clinical research, the question arises as to if and how radiology can contribute to today’s forensic examination of head and brain injuries. The in situ postmortem evaluation of the head and brain using CT and MRI has rarely been performed in the past. Only in recent years an increasing number of study groups started to belabor classical forensic topics such as traumatic alterations of the brain stem or cranial gunshot injuries by performing postmortem CT and MRI scans [10–27]. Yen et al. published the in situ neuroimaging results from 57 postmortem cases [28]. Apart from these studies, examinations of the formalin-fixed postmortem brain with the focus on neurotraumatology and neuropathology have been provided by several authors [29–33]. Furthermore, and also of some forensic interest, postmortem imaging of the fetal brain has become increasingly relevant for the clinical assessment of fetal brain alterations and injuries [34–39]. The following overview about the forensically most relevant neurotraumatologic and neuropathologic alterations and their radiological appearance is based on and according to current standard neuroradiological literature [8,40(and references therein)] and on seven years of research (2000 to *
Figure Captions: Not all findings that are relevant for forensic practice have so far been subject to postmortem imaging. To demonstrate the principal potential of the imaging methods also in these findings, they are shown in living patients in the following. Accordingly, Figure D3.3.1, Figure D3.3.5, Figure D3.3.24, Figure D3.3.29, Figure D3.3.30, Figure D3.3.31, Figure D3.3.41, Figure D3.3.42, Figure D3.3.43, Figure D3.3.44, Figure D3.3.46, Figure D3.3.47, Figure D3.3.48, Figure D3.3.49, Figure D3.3.50, and Figure D3.3.57 are from clinical examinations of living persons.
Incident-Specific Cases
2007) from the forensic postmortem imaging project—the Virtopsy Project—at the University of Bern in Switzerland. For a number of findings no postmortem imaging data exist up to date (e.g., venous sinus thrombosis, hypertensive intracerebral hematoma, degenerative diseases). According to the experience from routine autopsies, these issues are, however, relevant for future forensic radiological examinations, as they might cause sudden and unexpected death. Regarding findings where postmortem imaging knowledge is still lacking, the information provided in this chapter is based on clinical imaging experience only. According to the data from the virtopsy postmortem CT and MRI examinations, it is likely that findings seen in routine clinical examinations will also be detectable using postmortem imaging. The terms isodense, hyperdense, or hypodense in CT imaging and isointense, hyperintense, or hypointense in MRI define the density (CT) or signal intensity (MRI), respectively, in relation to unaffected brain parenchyma. For more detailed information on the radiological appearance of findings and diagnostics, the complementary lecture of specific neuroradiological literature is recommended.
D3.3.2 TECHNICAL ASPECTS For the forensic postmortem examination of cranial injuries and pathology, the usage of both imaging technologies CT and MRI is recommended. CT offers an excellent demonstration of bone findings and air inclusions, and the radiologists are well used to the method when it comes to detecting intracranial trauma findings. Even though requiring more effort and time for the examination itself, MRI has the advantage of being well suited for the diagnosis of soft-tissue lesions and of displaying some relevant additional diagnoses (e.g., diffusion processes in the brain tissues, age staging of hematomas) that cannot be obtained with CT. Furthermore, the limitations that widely prohibit the application of MRI in living acute-trauma patients are not relevant in postmortem situations, and the scan time is not as restricted as in clinical settings. Based on know-how from the Virtopsy Project [10,28], the cranial scans can be performed at least up to five days after death (when the
273
body has been stored at room temperature or below) without a relevant impairment of the image quality when compared with examinations of living persons.
D3.3.2.1 Postmortem In Situ Cranial CT Using modern multislice scanning technology, CT allows obtaining the radiological data of the head and brain within a few minutes and with excellent resolution. For postmortem examinations, no special preparation procedures apart from packaging of the body are necessary, and the same scanning protocols as those for the living can be used. For obtaining an optimized basis for high-resolution, three-dimensional image reconstruction, a small section thickness (e.g., 1.25 mm), a field of view of 250 mm or smaller, and a matrix of at least 512 r 512 should be used.
D3.3.2.2 Postmortem In Situ Cranial MRI As usual in clinical practice, it is recommended that a head coil be used for the postmortem examination. Offering a better signal-to-noise ratio, the head coil can easily be applied even when the body is packed in a supine position in a body bag. However, it is necessary to control the correct positioning of the head before starting the scan. The initial scout images will help in providing an overview about the position of the head and in finding the optimized orientation (Figure D3.3.1). If the head is intact and the anatomy is not severely altered, axial, coronal, and sagittal planes should be set in the same way as in clinical radiology [40]; this helps the radiologist to find orientation easier when analyzing the images. In each case it is necessary to perform the scan using at least two different orientations and two different weightings. According to the data from the Virtopsy Project on a 1.5 Tesla system (Signa, GE Medical Systems), a typical postmortem MR imaging protocol could have parameters as follows: T1-weighted spin-echo sequences (repetition time [TR] 400–600 ms, echo time [TE] 10–20 ms, matrix 256 r 192–224, field of view [FOV] 220–260 mm, slice thickness
FIGURE D3.3.1 The scout images that are issued at the start of each scan allow detailed planning and anatomical positioning (here shown in the coronal and sagittal plane with 20 axial slices each).
© 2009 by Taylor & Francis Group, LLC
274
4 mm, spacing 1–2 mm, 2 number of excitations [NEX], 22–44 slices depending on the slice direction and volume); T2-weighted sequences (TR 3500–5000 ms, TE 15–105 ms, matrix 256 r 192–224, FOV 24 cm, slice thickness 4 mm, spacing 1 mm, 2 NEX, 56–88 slices), and T1-weighted gradient-recalled echo sequences (TR 280–340 ms, TE 8–20 ms, matrix 256 r 224, FOV 240 mm, slice thickness 4 mm, spacing 1 mm, 2 NEX; 18–42 slices). The imaging protocols must, however, be adapted according to the local hardware and software of the used scanning unit, and the postmortem decreased body temperature must be taken into account. D3.3.2.3 Autopsy If autopsy is conducted to compare imaging and neuropathologic findings, it should be performed within a few hours after imaging to prevent bias from postmortem artifact. The autopsy technique should be adapted to the imaging protocols as far as possible (e.g., choosing the direction of brain slices in analogy to the scanning planes) to allow an optimized neuropathologic comparison and photographic documentation. This will, however, be difficult as one has to decide for one plane at autopsy, and the examiner will rarely be successful obtaining exactly the same plane as the one displayed at imaging (Figure D3.3.2). It can be advantageous to perform the neuropathologic examination on the formalin-fixed brain, even though fixation will also cause some artifact and thereby limit the direct comparability of autopsy and imaging findings. Histological sampling is needed for special investigations such as nerve-fiber alterations or small hemorrhages. D3.3.2.4 Image Reading and Forensic Expertise It is highly recommended that the reading of the radiological images is performed by experienced neuroradiologists. The description of the findings should contain their exact
The Virtopsy Approach
localization, extent (size and solitary vs. multiple lesions), radiological appearance (e.g., signal intensity, typical signs), and accompanying findings such as signs of edema or increased brain pressure. In routine casework, the forensic expert can provide the data from autopsy, if performed, and can add special forensic issues. It is the forensic pathologist who does the final forensic interpretation of the radiological findings and implements them into the final statements. In court, the images should be presented in a way that they are understandable also to nonmedical experts and laities. For this purpose, arrows or frames may be used to highlight the relevant findings (Figure D3.3.3), and the exact localization and plane of the image slices has to be explained. D3.3.2.5 Artifact and Limitations One general limitation of using imaging technologies for forensic and also clinical purposes is the problem of artifacts. In postmortem situations, artifact is often caused by the packaging or clothing of the body (e.g., metal parts or other dense materials). In forensic cases, artifact may also result from the incident itself (e.g., bullet particles, various other materials) (Figure D3.3.4). For the radiological examination it is therefore often advantageous to perform CT first and MRI not before the exclusion of larger metallic foreign parts. Concerning technical artifacts such as chemical shift or susceptibility artifacts (Figure D3.3.5) at MRI, the radiologists have the knowledge to detect these correctly and to distinguish them from pathologic findings. Obviously and in contrast to clinical radiology, movement artifacts play no role when performing the scans postmortem. Circulation standstill can, however, be another limitation: There is a lack of applying special techniques such as angiography (see Chapter D6). In bodies that present in a severely injured or decomposed status, it can be difficult to find orientation for defining the scanning planes (Figure D3.3.6). Another fact that may cause problems at the image reading is the postmortem status of the
FIGURE D3.3.2 Comparison of forensic neuropathology and imaging (in this case CT): It is difficult to obtain at autopsy the same correlative axial plane to imaging. CT with its excellent postprocessing possibilities, however, allows viewing the imaging data in all planes and directions, thus offering improved correlation between autopsy and imaging.
© 2009 by Taylor & Francis Group, LLC
Incident-Specific Cases
275
FIGURE D3.3.3 Arrows and frames are useful for explanation and demonstration of the relevant findings in court or in the classroom. This example shows diffusion tensors that are overlaid on sagittal T2-weighted MR images in a case who suffered blunt trauma of the brain stem. The white-matter tracts show disorganization at the same location that hemorrhage is seen in the autopsy correlate. The frames help to find orientation and to highlight the relevant findings.
brains, causing generally reduced gray–white matter distinction and being associated with alterations due to, for example, inner livores or decomposition processes. When performing reconstructions from the CT data, special attention has to be directed toward smoothing artifacts that might, for example, cover small fracture lines (Figure D3.3.7). Furthermore, the limited image resolution of CT and MRI itself restricts the diagnosis of very small findings (Figure D3.3.8).
FIGURE D3.3.4 Typical stripe artifacts due to metal particles are shown on this CT image from a suicidal gunshot case.
© 2009 by Taylor & Francis Group, LLC
Finally, one limitation that will be overcome in the future is the limited experience clinical radiologists have with examining forensic cases. The radiological appearance of special postmortem findings such as livores is unknown to most radiologists, and in some cases the widely altered anatomy leads to confusion. Radiologists also have to get used to some “new” questions such as forensic reconstruction based A
B
FIGURE D3.3.5 Axial and coronal T1-weighted (A) and echo-planar DWI images (B). On the DWI images, susceptibility artifacts are close to the skull base that create distortions that are absent on the T1 images.
276
The Virtopsy Approach
axis it is, however, extremely relevant to analyze not only the inner but also the superficial injuries at the localization of impact. When the impact axis is determined radiologically and is marked for presentation in court or the classroom, the analysis of eventual scalp lesions at the image reading will help with the forensic interpretation and presentation (Figure D3.3.9, Figure D3.3.10, Figure D3.3.11, Figure D3.3.12, Figure D3.3.13). If deep laceration of the scalp is present, CT and MRI imaging will display the lesion that will also be visible in the 3D reconstructed CT data (Figure D3.3.14). Not surprisingly, and according to a recent study [28], MRI is generally better suited for the detection of soft-tissue lesions and temporal muscle hemorrhage, a finding that is frequently seen in cranial trauma cases. D3.3.3.2 Skull and Upper Cervical Vertebrae FIGURE D3.3.6 Gross anatomical alteration is observed in this cranial CT image from a man who had been run over by a car on a highway. The massive destruction of the anatomical structures complicates the radiological evaluation.
on the radiology findings or defining the status of decomposition using radiological criteria.
D3.3.3 EXTRACRANIAL TRAUMA: SCALP AND MUSCLE TISSUES, SKULL, UPPER CERVICAL VERTEBRAE D3.3.3.1 Scalp and Soft Tissues, Temporal Muscles As traumatic injuries of the scalp tissues are usually well seen at the forensic external examination of the body, imaging will play a minor role concerning the detection of these findings. Furthermore, the presence of susceptibility artifacts at the body surface limits the radiological evaluation of this region. For the forensic reconstruction of the impact
A
Well known from clinical and postmortem neuroimaging, especially CT, provides an excellent basis for the diagnostics of osseous lesions due to its 2D and 3D reconstruction possibilities. Not only will the presence and localization of a fracture be detected and displayed, but the forensic examiner also will get important information about the fracture system (Figure D3.3.15 and Figure D3.3.16) and other characteristics within a few seconds after the scan. Furthermore, imaging reveals injuries also in regions such as the upper neck that are difficult to access at autopsy (Figure D3.3.17 and Figure D3.3.18). Autopsy does not provide analytic possibilities that are comparable to the excellent overview obtained by CT, and if maceration techniques are applied the diagnostic process takes at the very least hours to days. Despite the undoubted value of CT imaging for fracture detection, the reader must pay specific attention to small fracture lines and fractures that are localized at the base of the skull, as these might be overseen [28]. In the 3D reconstructed CT data, he or she must be aware that smoothing artifacts can hide small fractures (Figure D3.3.7).
B
FIGURE D3.3.7 Smoothing artifacts reduce the visibility of the fracture lines in this 3D reformatted CT image (A, frame). In contrast to 3D CT, the fracture is clearly visible on the axial CT slice (B, arrow). Generally and as in clinical routine, the axial CT slices shall be used for the primary evaluation of skull fractures; the 3D volume-rendered data are specifically useful for the forensic evaluation of the fracture system and visualization purposes.
© 2009 by Taylor & Francis Group, LLC
Incident-Specific Cases
277
A
B
FIGURE D3.3.8 Image resolution of the radiological methods is still limited, thus making the diagnosis of very small findings difficult. This example of a pedestrian who was hit by a car shows small secondary traumatic hemorrhage in the central brain stem (A) that is not seen at 1.5T MRI (B).
D3.3.4 TRAUMATIC INTRACRANIAL INJURY
detected using imaging methods. MRI is generally more sensitive than CT concerning these traumatic changes, allowing, if present, the detection of typical multiple lesions that appear hyperintense in T2-weighted and hypointense in the hemorrhage-sensitive gradient-echo T2*-weighted sequences [8]. These lesions will be predominantly found in the corpus callosum, the fornix and basal ganglia, the internal capsule, as well as in the border region between gray and white matter substance and the upper brain stem. Similar “black dots” can be found in hemorrhages that occur due to hypertension or lacunar or gross infarction; in amyloid angiopathy; or in cases with multiple vascular malformations; these reflect the main differential diagnoses. In DAI cases, MRI additionally offers the possibility to diagnose pathologically altered diffusion processes
D3.3.4.1 Commotio and Contusio Cerebri Subject to the definition of the clinically familiar term commotio cerebri, this diagnosis has no pathomorphological correlate [40]. Findings are therefore not to be expected at imaging or at autopsy or histological examination. The morphological appearance of primary traumatic lesions associated with cerebral contusion is discussed following. D3.3.4.2 Diffuse Axonal Injury Although morphological signs of diffuse axonal injury (DAI) are accessible only by means of histology in a number of cases, there are some radiological criteria that might be
A
B
FIGURE D3.3.9 81-year-old male who fell backward and hit his head against the floor. He was able to walk home, where he complained of dizziness. Two hours later he rapidly lost consciousness and died the next day. At autopsy, a contusion of the scalp was seen in the occipital region that was consistent with the reported trauma (A). No laceration of the skin was present. The T2-weighted MR image (B) displays the T2-hyperintense hemorrhagic changes in the occipital subcutaneous-tissue layers (frame). Death had occurred following progressive subdural hemorrhage.
© 2009 by Taylor & Francis Group, LLC
278
The Virtopsy Approach
A
B
FIGURE D3.3.10 The axial T1-weighted MRI in this case of suicidal gunshot (A) shows the left-sided fracture hematoma (B, frame) without alterations in the superficial cutis, which excludes hemorrhage due to a direct contusion. Note the collapsed brain and the intracranial air (b). (a) Autopsy correlation.
A
B
FIGURE D3.3.11 55-year-old male with self-inflicted gunshot wound. On the right side of the skull there is a hematoma covering the skull bone (A). No injury of the subcutaneous tissues or the cutis was seen at autopsy also in this case (frame). T2-weighted MRI (B) reveals the finding (arrows) and confirms its indirect genesis by displaying no subcutaneous lesions above the hemorrhage.
A
B
FIGURE D3.3.12 40-year-old female bicyclist who was hit by a car from behind. In the autopsy, subdural hematoma was detected (A). This is seen on the right side on CT (B, white arrows). Additionally, extensive left-sided cutaneous and subcutaneous hematoma (frame) and laceration (blue arrows) are present. In contrast to the indirect hemorrhages shown in Figure D3.3.10 and Figure D3.3.11, this finding was caused by a direct impact to the head, causing subcutaneous contusion and laceration. Congruently, CT reveals hemorrhage and swelling in all scalp-tissue layers, and a small laceration is visible (blue arrow).
© 2009 by Taylor & Francis Group, LLC
Incident-Specific Cases
279
A
B
SAR
FIGURE D3.3.13 This newborn child died a few hours after vaginal delivery. Forceps had been applied to the child’s head when the birth process became increasingly complicated. Forensic autopsy took place weeks after the incident on the disinterred body that was in a surprisingly good state of preservation. Previous to the autopsy (A), MRI revealed distinct cephalohematoma (B) that was later confirmed by autopsy.
in the brain tissues by means of diffusion-weighted imaging (DWI). If DAI is present, the apparent diffusion coefficient (ADC) will be markedly reduced in the injured regions (Figure D3.3.19). As a significant general reduction of the ADC is also observed as a specific postmortem finding [27] (Figure D3.3.51), the value of reduced ADC as a sign of DAI in postmortem cases has to be clarified in future studies. Besides giving information about the diffusion status, DWI is suited to display disorganization of the nerve fibers themselves by means of diffusion tensor imaging and fiber tracking, thereby offering new diagnostic possibilities in nonhemorrhagic brain trauma (Figure D3.3.20) [40]. As DWI is a relative new field of MRI, many issues concerning this technology are subject to current clinical studies. A
CT imaging can be normal in DAI cases. When hemorrhagic lesions are present, they might be seen as hyperdense small roundish spots predominantly in the predilection areas. D3.3.4.3
Extra-Axial Hemorrhage
D3.3.4.3.1 Epidural Hematoma In the medical literature the value of CT and MRI concerning the detection of space-occupying epidural hemorrhage is almost equal [40]. Both technologies are well suited to depict epidural bleedings (Figure D3.3.21). Epidural hematoma is characterized on CT by a hyperdense to mixed hyperdense and hypodense biconvex space-occupying lesion in the extra-axial space that is, in most cases, accompanied by a fracture of the skull. The
B
FIGURE D3.3.14 35-year-old male who was found lying in his bed in a state of advanced decomposition (same case as Figure D3.3.54). Semisharp injury had been applied to his head with an ax. On the skull there were parallel skin lacerations on the right side (A) that were consistent with the injury mechanism. Underneath the scalp, extensive skull fractures were present (Figure D3.3.54). The findings at the level of the skull cutis can be seen on the three-dimensional CT soft-tissue reconstructions (B) even in this progressively decomposed state.
© 2009 by Taylor & Francis Group, LLC
280
The Virtopsy Approach
(A)
(B)
(C) FIGURE D3.3.15 Self-inflicted suicidal gunshot to the head in a 37-year-old male. There was a frontal point of entry and a posterior exit wound, with a horizontal bullet track through the brain. Four different fracture systems could be observed in this case: (A) A characteristic whole-shaped fracture system at the region of entry. (B) A burst fracture on the right side of the skull. (C) A burst fracture in the anterior skull base. (D) A flexion fracture system posterior to where the point of exit was located. The projectile was placed in the exit wound again, causing some metallic artifacts that were visible especially on the 2D images. The three-dimensional CT reconstructions allow differentiation among the various fracture types. (A–D, left image) Autopsy correlation.
© 2009 by Taylor & Francis Group, LLC
Incident-Specific Cases
281
(D) FIGURE D3.3.15 (Continued)
brain tissue and the subarachnoid spaces below the hematoma are compressed, and midline shifting as well as signs of brain herniation can be present. At MRI, the dura mater might additionally be seen as a thin dark line between brain parenchyma and hemorrhage. The bleeding itself appears isointense to the cortex in most sequences. In postmortem examinations, blood segmentation is often seen within the hematoma, and small air bubbles are sometimes visible. According to results from a neurotraumatology study performed in the context of the Virtopsy Project (unpublished data), imaging mostly fails at detecting epidural blood layers smaller than 3 mm. If the epidural bleeding is situated in the infratentorial space, it is more difficult to see at both CT and MR imaging.
D3.3.4.3.2 Heat-Induced Epidural Hematoma Heat-induced epidural hematoma is a common postmortem finding in burned bodies (see Chapter D3.9.1). The radiological appearance of epidural heat hematoma is characterized by an isodense to hypodense, crumbly-appearing mass in the epidural space that shows numerous gas bubbles in CT (Figure D3.3.22). The skull bone frequently shows characteristic defects, and the external lamina is partially missing. In three-dimensional reconstructions from the CT data the epidural hemorrhagic layer with its typical crumbly structure can be seen. MRI is also suited for the detection of this specific finding.
FIGURE D3.3.16 32-year-old male who had a frontal collision with a truck while driving a car (same case as Figure D3.3.28). The driver cabin was compressed, and the man died immediately. Autopsy revealed a hinge-form fracture of the skull base (left) that is also visible on the three-dimensional CT reconstructions (right). In these 3D images, hole-shaped anterior skull base artifacts are recognizable; probably due to the presence of blood (verified by autopsy) at the base of the brain, in both orbital cavities, and the sinuses, the skull base is not differentiated from the surrounding tissues and is thus removed automatically by the post-processing application.
© 2009 by Taylor & Francis Group, LLC
282
The Virtopsy Approach
compressed sulci small hypodense liquor spots can be seen. In a number of cases subdural hematoma is associated with other traumatic findings such as parenchyma lesions or subarachnoid hemorrhage. As with epidural bleedings, according to studies from the Virtopsy Group (unpublished data), the finding might be overseen at imaging when the blood layer is smaller than 3 mm or is localized in the infratentorial space. Chronic subdural hematoma is, at CT imaging, characterized by a homogenous (hypodense) density that sometimes shows hyperdense septal structures, calcifications, or a “separated” appearance or hematocrit level caused by intermittent rebleeding (Figure D3.3.24). Subdural liquor accumulation might also be present (hygroma). MRI allows a time-dependent characterization of subdural hemorrhage in clinical examinations (Table D3.3.1) [8]. In forensic settings where the person died shortly after the incident, most subdural bleedings will correspond to the clinical definition of hyperacute (less than 12 h) or acute (hours to few days). It is, however, unknown if the radiological behavior of intracranial hemorrhage in postmortem examinations corresponds exactly to the clinical findings; further research concerning this issue will have to be provided.
FIGURE D3.3.17 Special insight can be gained from the imaging methods when regions that are difficult to access at the autopsy are evaluated. This example is from a 20-year-old male who died immediately on site when he was hit by a car from behind while riding his bicycle. The 3D volume-rendered CT excellently displays atlanto-axial luxation (left), an injury that might be missed at the autopsy due to the difficult access to that region. Additional MRI (sagittal T2-weighted image) (right) reveals posterior displacement of the dens axis, medullary transection (arrow), and subarachnoid hemorrhage covering the anterior medulla oblongata and spinalis. In the radiological images the distance between the anterior arch and the odontoid process can easily be measured (dashed line).
D3.3.4.3.4 Subdural Hemorrhage in Shaken Baby Syndrome Following shaking, besides the characteristic subdural bleeding CT and MRI might show signs of distinct brain edema and increased brain pressure as well as subarachnoid, intracerebral, and retinal hemorrhage. In some cases additional extracranial injuries (fracture screening) are also detected. If the acute phase after the incident has been survived, chronic subdural hematoma, hygroma, and atrophy might be present at the radiological examination. Concerning the radiological diagnosis of subdural hemorrhage in babies and young children, special attention must
D3.3.4.3.3 Subdural Hematoma At CT, hyperacute subdural hemorrhage is mostly seen as a hypodense blood layer. Acute cerebral subdural hemorrhage appears as a hyperdense or a combination of hyperdense and hypodense and sometimes inhomogeneous layer of blood covering one hemisphere (or in rare cases both) (Figure D3.3.23) (Table D3.3.1). Separation of blood components (i.e., fluid–fluid levels) might be present. The sulci are displaced from the skull toward the center, and in the
A
B
C
D
FIGURE D3.3.18 In this case of a 32-year-old male who died in a motor vehicle accident, CT and MRI revealed the neck findings clearly superior to the autopsy, providing an excellent overview of the region and the sustained injuries. The 2D reformatted sagittal CT scan (A) shows the fracture of the odontoid. The corresponding 3D volume-rendered CT (B) also clearly displays the osseous injury. Postmortem MRI provides information about the soft tissues and spinal cord: As seen in the sagittal T2-weighted fast spin-echo (FSE) image (C), besides the odontoid fracture there is a partial ablation of the posterior longitudinal ligament (arrows), while the brain stem and spinal cord appear intact. This was confirmed at the autopsy, where the brain stem and spinal cord showed no signs of traumatic injury (D) after formalin fixation).
© 2009 by Taylor & Francis Group, LLC
Incident-Specific Cases
283
FIGURE D3.3.19 DTI of a case with posttraumatic shearing injuries following a fall from height. The subcortical hyperintensities (frames) that correspond to reduction of the ADC represent subcortical diffuse axonal lesions. A
B
D
C
E
FIGURE D3.3.20 An MRI application of specific forensic interest is DWI with its possibilities of tracking fibers in the brain white matter. Fiber tracking based on DTI is currently subject to numerous clinical studies with regard to its potential application in tumor patients for surgery planning and the assessment of neurodegenerative disorders. In forensic medicine, DTI could offer the first step into a non-hemorrhage-based neurotraumatology, as it offers a “direct” detection and display of nerve fiber lesions. The example shown demonstrates sagittal diffusion tensors overlaid on T2-weighted MR images in three cases (the principal eigenvector of the diffusion tensor scaled by the fractional anisotropy is represented by its projection onto the sagittal plane using red dashes and by its through-plane component using dots: yellow dot, strong component; blue dot, intermediate component; missing dot, weak component): (A) 49-year-old male who crashed while hang-gliding, suffering blunt injury of the distal brain stem. (B) Male who died from cardiac standstill. (C) Living healthy volunteer. The injury case (A) shows a traumatic disorganization of the brain-stem white-matter tracts (frame). The cardiac arrest case presents with an overall slight wavy fiber configuration, a finding that was observed in all postmortem cases. The vital control case (right) has a regular fiber orientation. (D,E): The morphological correlates to (a) demonstrating hemorrhage of the brain stem at the gross (D) and histopathological (e) examination (hematoxylin and eosin staining).
© 2009 by Taylor & Francis Group, LLC
284
The Virtopsy Approach
A
B
C
D
FIGURE D3.3.21 Traumatic left parietal epidural hemorrhage with typical biconcave configuration in a case who suffered accidental blunt force to the head when he was hit by a falling tree (same case as in Figure D3.3.26). The CT shows clearly the adjacent parietal fracture line (B, arrow). Note also the contralateral parietal and frontal fractures, the subdural and subarachnoid hemorrhages, and frontal hematosinus: (A) Autopsy. (B) CT. (C) T1-weighted MRI. (D) T2-weighted MRI.
be given to the differential diagnosis between hygroma and hemorrhage. From our own case samples, several babies who initially survived shaking were misdiagnosed as hygroma at the clinical CT examination. In all these cases, shaking had not been reported as the injury-causing mechanism by the parents initially. After death a few hours or days after admission to the hospital, at autopsy acute subdural hemorrhage following forceful shaking was found.
depict subarachnoid hemorrhage superior to postmortem CT. This was obviously due to the higher potential of MRI to demonstrate the finding despite the presence of cerebral edema or expansive epidural and subdural hemorrhages. However, at both CT and MRI, there was a relatively high number of falsepositive and false-negative diagnoses of subarachnoid hemorrhage, which were mainly due to artifact influence (e.g., metal artifacts, artifacts due to pneumencephalon) and in some cases were caused by misinterpretation.
D3.3.4.3.5 Subarachnoid Hemorrhage With few exceptions, subarachnoid hemorrhage is accompanied by other traumatic lesions such as cerebral contusions, subdural hemorrhage (Figure D3.3.25), soft-tissue injury of the scalp, or skull fracture. At CT imaging, subarachnoid bleeding appears as hyperdense areas in the sulci or cisternae, mostly localized close to other traumatic findings. Using MRI, the hemorrhagic sites show high signal intensity in FLAIR sequences, whereas they appear isointense to the brain tissue in T1- and T2-weighted sequences. For this reason CT seems to be better suited for the detection of the finding. In a current study from the Virtopsy Project (unpublished data), postmortem MRI was found to
© 2009 by Taylor & Francis Group, LLC
D3.3.4.4 Intra-Axial Hemorrhage D3.3.4.4.1 Cortical Contusions Cerebral contusions are frequently seen findings in forensic cases who suffered head trauma. Their detection is of specific forensic interest, as they help to determine the impact direction and therefore the sequence of events (Figure D3.3.36). MRI is generally superior to CT for the evaluation of cortical hemorrhages. From a forensic view, MRI has the additional advantage of better displaying accompanying injuries of the scalp soft tissues than CT, which can be useful for forensic
Incident-Specific Cases
285
A
C
B
D
E
FIGURE D3.3.22 Heat-induced, frontal-accentuated epidural hematoma with typical crumbly appearance of the coagulated epidural blood. Note also the almost ubiquitary characteristic defects of the external tabula better seen on CT (A), the pneumocephalon as well as parenchymal shrinkage due to heat exposure: (A) Autopsy. (B) CT. (C) T1 weighted. (D) T2 weighted. (E) Gradient echo. A
C
B
D
E
FIGURE D3.3.23 Traumatic right-hemispheric acute subdural hematoma following blunt force trauma (same case as in Figure D3.3.32, Figure D3.3.34, and Figure D3.3.53) with typical concave configuration, ispsilateral parenchymal compression, and contralateral midline shifting: (A) Autopsy. (B) CT. (C) T1 weighted. (D) T2 weighted. (E) Gradient echo.
© 2009 by Taylor & Francis Group, LLC
286
The Virtopsy Approach
TABLE D3.3.1 Age-Dependent Appearance of Subdural Hemorrhage (SDH) Hyperacute SDH Acute SDH (few hours to approx. 3 days)
Subacute SDH (3 days to approx. 2 weeks) Chronic SDH (older than 2 weeks)
CT Mostly hypodense Hyperdense Hyper- and hypodense
MRI
T1: Iso- to hypointense T2: Hyperintense Gradient Echo (GRASS): Profoundly hypointense Iso- to hyperdense T1: Hyperintense T2: Hypointense Homogenous hypodensity, or T1: Hyper- or hypointense Inhomogenous density (blood segmentation, septae, calcification) T2: Hyperintense
Note: The MR appearance of hemorrhage depends on the MRI techniques and imaging parameters.
reconstruction. On the other hand, CT allows accompanying skull and skull base fractures to be diagnosed that might escape MRI (e.g., orbital fractures at the contrecoup site). Brain CT findings can be unsuspicious even when cortical contusions are found at autopsy. This is the case especially when the person died shortly after the incident. Furthermore, the generally reduced gray–white matter distinction that is observed in postmortem cases can cause diagnostic problems. However, the hemorrhagic lesions can occur as multiple hyperdensities at CT, which might be accompanied by hypodense edema of the surrounding cortex (Figure D3.3.26). MRI usually depicts the findings as multiple hyperintense or hypointense areas surrounded by T2-hyperintense cortical edema (Figure D3.3.27). As with other hemorrhage types, there is a time dependency of the imaging behavior at MRI as well as a dependency from the imaging protocols that are used. The data from the Virtopsy Project (unpublished material) have shown that both CT and MRI regularly fail in the detection of intracerebral hemorrhagic lesions smaller than about 3 mm (Figure D3.3.28), especially when there are only few or solitary lesions without accompanying edema. The same
has been observed regarding old cortical contusion hemorrhages (“plaques jaunes,” localized atrophy) in patients who had survived a traumatic incident. When taking into account the rapid technical developments in clinical radiology, this lack of depicting small and solitary but forensically highly relevant hemorrhages possibly will be overcome when using 3 T or higher magnet MRI scanners and specially adapted imaging protocols. As the depiction of small findings is also of clinical interest, it is addressed in current clinical MRI studies. D3.3.4.4.2 Intracerebral Hemorrhage As with small contusion hemorrhages, traumatic intracerebral bleedings smaller than about 3 mm mostly escape radiological detection at 1.5T MRI and CT. Larger hematomas show similar criteria to the aforementioned cortical findings with characteristic time-dependent signal changes, being initially hyperdense at CT imaging and showing a mixed signal intensity accompanied by T2-hyperintense edema at MRI (Figure D3.3.29, Figure D3.3.30, Figure D3.3.31, Figure D3.3.32, and Figure D3.3.33). The hemorrhage shows signs of separated blood layers in some cases. The brain might present with midline shifting and increased brain pressure (Figure D3.3.29 and Figure D3.3.30). D3.3.4.4.3 Ventricular Hemorrhage In forensic examinations the ventricular system is often hemorrhagic following major head trauma. Usually this is detected without difficulties at both CT and MRI. CT reveals hyperdensities in the ventricular system and the plexus choroideus region (Figure D3.3.26), and using MRI the finding is characterized by imaging protocol- and timedependent hemorrhage characteristics as described above (see section D3.3.4.3.3). Radiological signs of blood separation occur frequently. D3.3.4.4.4 Brain-Stem Hemorrhage
FIGURE D3.3.24 The CT images show chronic left-sided subdural hematoma with characteristic blood sedimentation and hyperdense septal structures in a clinical patient.
© 2009 by Taylor & Francis Group, LLC
Not only the principal detection of brain-stem hemorrhage but also the differentiation between primary and secondary traumatic brain-stem findings is important for forensic diagnosis
Incident-Specific Cases
287
A
C
B
D
E
FIGURE D3.3.25 Traumatic combined ubiquitary subdural and subarachnoid hemorrhage following blunt force impact in a 75-year-old male who was hit by a car while riding his bicycle (same case as in Figure D3.3.33). Note the fluid-fluid level on the left-hemispheric subdural hematoma that is seen especially well at MRI (C–E): (A) Autopsy. (B) CT. (C) T1 weighted. (D) T2 weighted. (E) Gradient echo.
and interpretation (Figure D3.3.34 and Figure D3.3.35). Even though such a differentiation is not familiar in clinical radiology, it seems possible based on analysis of the localization and morphological characteristics of the hemorrhagic findings. For distinguishing primary from secondary brain-stem injuries, clinical aspects regarding the survival time and A
B
FIGURE D3.3.26 Blunt force trauma to the head (A) in a man who was hit by a tree and died immediately on site (same case as in Figure D3.3.21). Besides cortical contusion hemorrhage in the frontal and left-sided brain regions, CT reveals intraventricular hemorrhage on the right side (B, frame). The latter finding is almost not seen at autopsy (A), which is obviously due to the fact that the blood leaks out when the Flechsig’s cut is applied.
© 2009 by Taylor & Francis Group, LLC
FIGURE D3.3.27 30-year-old male who committed gunshot suicide in his car. Death occurred following a central breathing paralysis and gas embolism. There was a direct wound to the brain stem with multiple hemorrhagic lesions due to hydrodynamic pressure waves (left). Despite the fact that most of the small hemorrhages were below the capacity for resolution of the imaging scans, the coronal T2-weighted MRI shows hyperintense and hypointense regions and spots in the brain stem (right). Additionally, signs of edema and massively increased brain pressure are present with tonsillar herniation (right, arrows) as well as uncal herniation (right, frame) on the left side.
288
A
The Virtopsy Approach
B
FIGURE D3.3.28 Same case as in Figure D3.3.16 (32-year-old car driver who died on site following a frontal collision with a truck). At the frontal poles punctuate hemorrhagic contusions are visible at the direct impact site (A). These findings are not seen on the CT, probably due to them being smaller than what the resolution capacity of the scanner is able to detect. MRI also did not reveal the hemorrhages. 3T MRI might hopefully overcome this limitation, as such findings are important for forensic reconstruction.
the clinical presentation of the patient must also be considered. In general, primary traumatic bleedings predominantly reflect small hemorrhages localized in the lateral parts of the brain stem or the cerebellar peduncles and tectum, as well as in the subependymal tissues of the third and fourth ventricle and the aqueduct. Secondary findings that occur due to increased brain pressure are mostly found in the central brain stem or rostral toward the midbrain and in the paramedian peduncles ;41=. Traumatic hemorrhage of the brain stem is usually associated with other cranial traumatic injuries (e.g., contusion hemorrhages or lacerations of the brain, skull or skull base fractures) that allow differential diagnosis from nontraumatic hemorrhages. Using CT and MRI as diagnostic methods, traumatic brain-stem hemorrhage might remain undetected due to current technical limitations concerning image resolution. Especially if the findings are smaller than 2–3 mm and singular, they will escape detection at radiology (Figure D3.3.8).
A recent study [42], however, demonstrated that special MRI applications have a promising potential to expand the diagnostic spectrum in forensic neurotraumatology: Using diffusion tensor imaging (DTI), it was possible to display traumatic disorganization and rupture of nerve fibers in the brain stem, offering diagnosis based on a “direct” examination of the fiber structures independently from the extent of accompanying hemorrhage (Figure D3.3.3 and Figure D3.3.20). The application of diffusion-weighted MRI is central in current neuroradiological studies, and in forensic pathology this technique may open the door toward a “direct” neurotraumatology based on fiber analysis. Today, the diagnosis of neurotraumatic alterations is mainly based on the “indirect” evaluation of hemorrhage. The radiological detection of brain-stem contusion and small lacerations is therefore based on the presence and extent of hemorrhagic lesions. In the Virtopsy Series, disruption of the neural structures in larger brain-stem lacerations was detected clearly at imaging. MRI was superior to CT, and imaging was generally superior to autopsy in these cases, offering an excellent overview in the neck region that is traditionally difficult to access at autopsy (Figure D3.3.17 and Figure D3.3.18) [19].
D3.3.5 TRAUMATIC INJURIES: BLUNT FORCE TRAUMA Cranial injury due to blunt force trauma as a major cause of severe disablement or death is an issue that has great relevance for the forensic examiner [43]. All forms of major and minor intracranial injuries might be caused by the various forms of blunt traumata. In clinical imaging, CT is the method of choice, especially in the acute situation, while MRI plays a minor role due to its specific limitations concerning the examination of living persons (e.g., scan duration, restrictions for intensive care patients). Therefore, the clinical experience with MR imaging of patients following acute head trauma is still quite low. However, postmortem forensic MRI examinations have shown that there is a specific potential of this method not only to display the “classical” neurotrauma
FIGURE D3.3.29 Hyperacute intracerebral hematoma: In the DWI image (left) an inhomogeneous mass is present that appears slightly hyperintense on T2-weighted images (center) but is isointense on T1-weighted images (right). The left ventricle is compressed, and midline shifting to the right side is visible.
© 2009 by Taylor & Francis Group, LLC
Incident-Specific Cases
289
A
FIGURE D3.3.30 Subacute intracerebral hematoma in the right striatum. There is a peripheral T1-hyperintense rim (left) and a central hypointensity on the T2 images (right).
findings but also to detect findings that might be overestimated or underestimated or even that remain undetected at autopsy ;19,28,42=. The main causes of cranial blunt force injuries are traffic incidents, falls from height (Figure D3.3.36), nonaccidental trauma, and sports-induced injuries. For the specific findings that are characteristic to the particular genesis of the cranial injuries, we refer to the specific forensic literature [43,44].
B
FIGURE D3.3.31 Chronic intracerebral hematoma: The finding shows cerebrospinal fluid (CSF) isointensity on both T1-weighted (A) and T2-weighted (B) images.
The typical radiological features that point toward child abuse and shaken baby syndrome are listed in [45].
D3.3.6 TRAUMATIC INJURIES: PENETRATING INJURIES D3.3.6.1 Gunshot Injuries The traumatology and morphology of gunshot injuries is described in detail in Chapter D3.5. In general and according to the Virtopsy data, CT and MRI seem to be well suited for the localization of bullet particles and many gunshot-induced
B A
C
D
E
FIGURE D3.3.32 Extensive secondary traumatic intracerebral hemorrhage in the right temporo-occipital region with surrounding edema and secondary brain-stem hemorrhage in a man who fell and survived for hours (same case as in Figure D3.3.23, Figure D3.3.34, and Figure D3.3.53). Note the adjacent subdural and subarachnoid hemorrhage: (A) Autopsy. (B) CT. (C) T1 weighted. (D) T2 weighted. (E) Gradient echo.
© 2009 by Taylor & Francis Group, LLC
290
The Virtopsy Approach
A
on the weapon and projectiles used, the injuries can cause massive destruction of the head and brain and therefore cause problems for the investigating radiologist concerning the exact anatomical localization and characterization of the findings (Figure D3.3.37 and Figure D3.3.38).
B
D3.3.6.2 Penetrating Injuries: Others
FIGURE D3.3.33 75-year-old male bicyclist who was hit by a car (same case as in Figure D3.3.25). A small hemorrhagic lesion is visible in the left gyrus cinguli (A, arrow), also depicted in T1-weighted MRI (B, arrow). There is also extensive bilateral subarachnoid hemorrhage that is more pronounced on the left side, a finding congruent with the autopsy result.
injuries like hemorrhage, contusion, and laceration of the extracranial and intracranial tissues. Frequent accompanying findings such as pneumencephalon, gas embolism and intracranial bone fragments can also be detected at imaging. The reconstruction of the bullet path will in many cases be possible when not only the intracerebral but also the skull and softtissue injuries are taken into account. However, depending
Postmortem Imaging data from intracranial stab wounds, bite wounds, or other penetrating injuries such as splinter injuries following bomb attacks or explosions are still restricted to single cases. The clinical CT examination of surviving patients has shown that imaging is especially useful for the exact localization of foreign material inside the cranium. This might also improve future forensic reconstruction in these cases if the deceased are investigated using imaging methods. D3.3.6.3 Brain Laceration Many open-head injuries are associated with laceration of the brain tissue. As with other pathologies, very small lacerations of a few mm escape today’s imaging possibilities. For larger injuries, CT and MRI reveal the destroyed tissue as an irregular lesion where parts of the brain parenchyma can be missing (Figure D3.3.37 and Figure D3.3.38). In peripheral
B
A
C
D
E
FIGURE D3.3.34 The same case as in Figure D3.3.23, Figure D3.3.32, and Figure D3.3.53, showing secondary traumatic brain-stem hemorrhage: (A) Autopsy. (B) CT. (C) T1 weighted. (D) T2 weighted. (E) Gradient echo. Due to the dramatic increase of intracranial pressure, multiple hemorrhages in the pons and cerebellum occurred, which can be seen on the radiological images.
© 2009 by Taylor & Francis Group, LLC
Incident-Specific Cases
A
291
B
air and gas inclusion are penetrating blunt and sharp force, burns, and diving incidents. Independent from the embolism-causing factor, the findings are well depicted at imaging even when only small amounts of gas are present, and even a quantification of the amount of gas is possible using the imaging data [46]. The differential diagnosis from gas inclusion due to decomposition processes is possible by analyzing the injury-causing mechanism, accompanying injuries, and especially the localization of gas. In decomposition, gas will be found in a more or less equal distribution intravascular and extravascular in all tissue layers including the scalp and muscles (see Chapter D2.1). D3.3.7.2 Cerebral Fat Embolism
FIGURE D3.3.35 Primary traumatic hemorrhage of the spinal cord below the pons (A) in a female pedestrian hit by a car. The T2-weighted FSE MR images demonstrate the injury (B, frame).
lacerations, which reflect the majority of these findings, the cortex will be more affected. Accompanying hemorrhage and edema or near-situated skull or skull base defects are present in many cases.
Some specific findings associated with cerebral fat embolism have been described in the radiological literature, such as multiple small nonconfluent hyperintense intracerebral lesions larger than 2 mm on proton-density and T2-weighted MR images or multiple low-density areas at CT [47,48]. The postmortem appearance of fat embolism has not been evaluated in forensic cases until today. However, some clinical data emphasize the diagnostic potential of DWI regarding this finding. In several studies punctuate hyperintensities in diffusion-weighted MRI sequences were reported besides signs of cytotoxic edema [49,50].
D3.3.6.4 Decerebration
D3.3.7.3 Foreign Bodies
Partial or complete decerebration is a finding that is sometimes seen in postmortem forensic examinations. Unsurprisingly, the clinical examiner is not used to these findings, and problems with the image interpretation result from the gross anatomical alteration that is usually present (Figure D3.3.6, Figure D3.3.37, and Figure D3.3.38). However, distinct anatomical allocation is of minor forensic relevance in many of these cases, and a definition of the most relevant injuries will often be possible without major difficulties. For the detection of dura mater and tentorium ruptures, MRI is better suited than CT; the dura appears as a hypointense band structure that is dislocated from the skull bones.
Foreign bodies such as projectile parts or bone fragments as emboli in the cerebral blood vessels reflect a rare finding in forensic cases. If present, these might be detected at imaging even better than at autopsy due to the excellent multiplanar overview that is provided by CT and MRI and its 3D reconstruction possibilities. Clinical reports have shown that not only the embolus but also the complications caused by it can be detected [51,52].
D3.3.8 NONTRAUMATIC INJURY D3.3.8.1 Cerebrovascular Diseases D3.3.8.1.1 Cerebral Insult
D3.3.7 EMBOLISM: GAS, FAT, FOREIGN BODIES D3.3.7.1 Gas Embolism Since radiological examinations have been performed frequently for forensic purposes, the diagnosis of gas embolism has noticeably increased. Within the Virtopsy Project all deceased who had suffered a gunshot to the head showed cerebral (and extracerebral) gas embolism that was especially visible in the CT data (Figure D3.3.39). CT turned out to be far superior to the autopsy concerning the detection of intracranial gas. This is not surprising, as it is often impossible to distinguish iatrogenic gas intrusion (caused by the opening of the cranial cavity) from preexisting gas embolism at autopsy. Other traumatic events that are frequently associated with
© 2009 by Taylor & Francis Group, LLC
At autopsy, the diagnosis of cerebral insult may be difficult. Small or hyperacute insults can easily be overlooked with the naked eye if histological sampling and examination are not performed. MRI is especially advantageous over autopsy concerning acute cerebral insult, offering an early and reliable diagnostic tool for the detection of hypoxic brain lesions in vivo [8,40,53,54] (Figure D3.3.40). MRI, with its new application DWI, displays ischemic lesions shortly after the onset of insult as a hyperintense region with reduced diffusibility (Figure D3.3.41). In the T2-weighted MR images, the hypoxic tissues can appear as hyperintense areas. The gray– white differentiation is often reduced, and signs of brain swelling can be present [8,40]. In native CT, a hypodense region may be seen as well as a hyperdense blood vessel that
292
The Virtopsy Approach
A
C
B
E
D
F
G
H
FIGURE D3.3.36 51-year-old male who fell from a roof while working. He immediately became unconscious and died a few hours later. On the left side of the head a small traumatic wound with hemorrhage into the cutis and subcutis was seen at the external examination (A). On CT the small laceration itself cannot be detected (arrow); however, massive swelling and subcutaneous hemorrhage in the impact region are well depicted (B). A left-sided burst fracture system is present (D, frame) as well as temporal left cortical contusion and laceration at the fracture site (C, autopsy; D, frame, CT) and a typical contrecoup lesion in the right frontal lobe (D, yellow arrows). A small fracture of the right-sided orbital roof is also found at CT (E,F). T2-weighted MRI reveals anterior subarachnoid hemorrhage (G, autopsy; H, MRI). Based on the localization of the various lesions, the impact axis can easily be reconstructed (D, white arrow). As an additional finding, there is a small falcine calcified meningioma (D, blue arrow).
is sometimes present, and often the differentiation between gray and white matter is reduced. First, postmortem imaging data have shown that cerebral insult is also detected well at postmortem MRI, with a similar appearance as in living persons. However, the influence of postmortem changes (e.g.,
© 2009 by Taylor & Francis Group, LLC
temperature decline, changes in diffusion processes) on the imaging appearance is not well known to date and will be subject to future research. Due to the suspended circulation, in postmortem examinations there is a lack of performing perfusion or contrast-
Incident-Specific Cases
293
A
B
case of a nontraumatic genesis, the hematoma will be solitary, without other accompanying traumatic injuries but with a local surrounding edema. D3.3.8.1.3 Nontraumatic Subarachnoid Hemorrhage
FIGURE D3.3.37 Brain laceration and partial decerebration following a suicidal gunshot to the mouth. T2-weighted MRI shows extensive laceration and partial decerebration predominantly in the frontal brain region. The intracranial anatomy is severely altered due to the injury: (A) Sagittal slice. (B) Axial slice.
In living persons with aneurysmatic subarachnoid hemorrhage, CT reveals hyperintensities in the basal subarachnoid spaces (Figure D3.3.43). MRI is less suited for the detection of this finding as the hemorrhage often appears isointense to the brain parenchyma. According to the radiological literature, FLAIR sequences are the most sensitive at MRI, demonstrating the bleeding with high signal intensity. Concerning the detection of the aneurysm that caused subarachnoid hemorrhage, autopsy is far superior to forensic imaging due to the lack of adequate postmortem angiographic methods. D3.3.8.1.4 Thrombosis of the Venous Sinuses
enhanced imaging. These techniques are applied in daily clinical practice when examining patients with suspected cerebral infarction or vessel disease. There are, however, attempts to develop techniques for a postmortem application of contrast media (see Chapter D6), and the results of these studies are promising. Future forensic radiological research will define the value and possible applications for these technologies. D3.3.8.1.2 Hypertensive Intracerebral Bleeding The differential diagnosis of hypertensive intracerebral bleeding versus trauma-induced intracerebral hemorrhage is important in forensic examinations. Acute hypertonic intracerebral hemorrhage is often extensive and usually localized in the basal ganglia and the central brain regions or the pons and cerebellum. Usually hypertonic bleeding can be diagnosed at both CT (hyperdense area) and MRI (sequence and time-dependent) without difficulties (Figure D3.3.42). In
A
The main problem of postmortem CT and MRI concerning the detection of thromboses of the venous sinuses is the arrest of blood circulation in the dead and the lack of performing contrast-enhanced examinations. According to clinical experience, in native CT the affected dural sinus in some cases appears hyperdense (Figure D3.3.44), and the brain parenchyma might show signs of small hemorrhages and edema. At MRI, the thrombus shows varying signal intensity: isointense in T1 and hypointense in T2 in acute thrombosis; hyperintense in T1 and T2 in the subacute phase. Venous infarction is characterized by swelling of the gyri and flattening of the sulci in the affected regions. Small cortical or subcortical hemorrhages and T2-hyperintense focal brain lesions can be present. D3.3.8.2 Neoplasm To give an overview of the main forms of intracranial tumors and their radiological appearance would go far beyond the
B
FIGURE D3.3.38 Partial decerebration following a suicidal gunshot in CT. The large bone defect corresponding to the entrance-wound region at the right side of the skull is easily seen (A). In the remaining brain tissue, bone and bullet particles are present besides air inclusions (B). Corresponding finding at the scene of death. One hemisphere had been thrown out from the head.
© 2009 by Taylor & Francis Group, LLC
294
The Virtopsy Approach
scope of this chapter. Furthermore, and in contrast to clinical neuroimaging, tumor alterations are of minor relevance in forensic casework. Even though own data from the Virtopsy Project (Figure D3.3.45) are still rare, there are extensive clinical data concerning the detection and classification of these alterations by CT and MRI. The neuroradiological literature gives detailed information about the radiological tumor characteristics [8]. D3.3.8.3 Inflammation To date, knowledge concerning forensic postmortem imaging of inflammation processes inside the cranium is rare. Based on clinical radiology it is likely that CT and MRI have a certain potential to depict inflammatory changes before autopsy. This could help the forensic examiner to plan and apply special examination techniques (i.e., liquor puncture, tissue sampling) before entirely opening the cranial cavity and therefore could help with preventing autopsy personnel from being contaminated with infectious material. In the following sections, the typical clinical radiological signs of native CT and MRI are described for the main forms of intracranial infection [8,40]. The application of contrast media is used for clinical examination in daily radiological routine; however, these techniques to date cannot be applied in postmortem situations.
FIGURE D3.3.39 Gas embolism following a suicidal gunshot to the head that had caused vessel lesions. The gas distribution in the transverse sinuses, the nasal sinuses, and the mastoid cells is seen perfectly in the 3D air structure volume-rendered CT image. The ruptured transverse sinuses (arrows) were considered as being the portal of air entry that caused air embolism of the heart. A
C
D3.3.8.3.1 Meningitis Imaging findings in acute meningitis can be unspecific, but radiological methods well display the complications of B
D
E
F
FIGURE D3.3.40 Posttraumatic ischemic bioccipital and left insular insult in a woman who fell down the stairs. The extent of the insult is depicted by MRI superior to the autopsy, especially in the T2-weighted (D) and diffusion (F) sequences where the ischemic region appears hyperintense. Note the left-hemispheric subdural hematoma: (A) Autopsy. (B) CT. (C) T1 weighted. (D) T2 weighted. (E) Gradient echo. (F) DWI.
© 2009 by Taylor & Francis Group, LLC
Incident-Specific Cases
A
295
B
C
D
FIGURE D3.3.41 Acute right-hemispheric stroke in a living patient. The hypodensity on CT (A) corresponds to hyperintensity on the DWI (B) with diminished ADC (C). The late CT (D) shows chronic hemispheric infarction.
inflammation (e.g., hydrocephalus, cerebrovascular complications). In native CT, the ventricles might be slightly enlarged, and the basal cisterns may be narrowed. MRI in some cases depicts the T1-isointense and T2-hyperintense exsudate in the sulci and cisterns (Figure D3.3.46). D3.3.8.3.2 Empyema Empyema is visible as a biconvex extra-axial attenuation of fluids that is mostly situated bifrontal in the epidural or
FIGURE D3.3.42 Typical appearance of hypertensive hematoma at DWI (B 1000); (A). (B) T1 weighted. (C) Photodensity weighted. (D) T2 weighted (MRI). All localized in the left-sided basal ganglia. © 2009 by Taylor & Francis Group, LLC
subdural space or both. The adjacent brain tissue can appear hyperintense in T2 at MRI. The differentiation from chronic subdural hemorrhage is possible by MRI, which displays the typical time-dependent blood characteristics when hemorrhage is present (See section D3.3.4.3.3). D3.3.8.3.3 Brain Abscess An abscess within the brain tissue is typically characterized by a hypodense subcortical lesion at CT with an irregular margin (which appears increasingly hyperintense when contrast media are applied). Time-dependent increase of local edema and a space-occupying lesion are often seen. MRI reveals an irregular, mixed hypointense and hyperintense (T1) or hyperintense (T2) lesion in the early stages. Later, the center of the lesion becomes hypointense in T1 with an isointense to
FIGURE D3.3.43 Aneurysmal subarachnoid hemorrhage in a clinical patient. The CT shows hyperdense blood in all subarachnoid spaces.
296
The Virtopsy Approach
FIGURE D3.3.44 Thrombosis of the superior sagittal sinus. The initial CT shows spontaneous hyperdensity of the sinus without contrast (top left) and a “delta sign” after injection (top, middle). The T2-weighted MRI shows hyperintensity of the superior sagittal sinus (top right). Phlebo-MRI depicts irregularities of the posterior portion of the superior sagittal sinus (bottom left), which appears also spontaneously hyperintense on the sagittal T1 image (bottom right).
slightly hyperintense peripheral “ring” (hyperintense center with hypointense “ring” in T2) (Figure D3.3.47). A capsule might be present and visible at MRI. DWI and MR spectroscopy allow the differential diagnosis between tumor and abscess in most cases: In contrast to most tumors, an abscess shows an increased signal intensity in the DWI images, and some specific markers can be found at spectroscopy (e.g., lactate, acetate).
DWI-hyperintense areas with diffuse margins (Figure D3.3.48). Signs of edema and hemorrhage might be present, and in some cases accompanying meningitis is seen. Depending on the infectious agent, the main localization of the encephalitic processes varies (e.g., central white matter in HIV encephalitis, gyrus cinguli and temporal lobes in herpes simplex type I infections). CT imaging depicts no pathologic findings in a number of cases, especially in the early acute phase. Later, hyperdensities in the affected regions and signs of edema and bleeding can be found.
D3.3.8.3.4 Encephalitis
D3.3.8.3.5 Tuberculosis of the Central Nervous System
MRI is more sensitive than CT in the detection of encephalitic alterations. MRI shows large T1-hypointense and T2-and
A
Tuberculosis is observed regularly at autopsy, and in a number of cases it is not known to the examiners before the
B
FIGURE D3.3.45 Falx meningeoma in a man who died after a fall (same case as in Figure D3.3.36): (A) Autopsy finding. (B) T2-weighted MRI. The finding had no relevance concerning the incident or the cause of death.
© 2009 by Taylor & Francis Group, LLC
Incident-Specific Cases
A
297
B
FIGURE D3.3.46 Acute meningitis is characterized by strong diffuse contrast enhancement of the meninges at both CT and MRI (A,B, coronal T2-weighted MRI).
examination. Imaging could therefore give important clues before the body is opened. Meningitis in the basal areas of the brain is typical for tuberculosis, especially when signs of tuberculosis are also found in other organ systems, and hydrocephalus and infarction are regularly seen as accompanying findings. Tuberculosis-induced granuloma is present at CT imaging as a roundish hyperdense or hypodense lesion, sometimes with accompanying edema. MRI reveals T1-hypointense and T2-hyperintense lesions (noncaseous granuloma) or slightly T1-hyperintense and T2-hypointense lesions (caseous granuloma), respectively. According to clinical experience, the differential diagnosis between tuberculosis associated alterations and other findings (e.g., tumor, abscess, hemorrhage) is often difficult. D3.3.8.4 Hereditary Malformation Although postmortem imaging within the Virtopsy Project did not include malformation cases, clinical radiological praxis has shown that malformations are generally diagnosed well at MRI and CT. Due to their ability to display the intracranial structures in all planes and directions without being
A
B
FIGURE D3.3.47 Frontal right-sided brain abcess. At MRI, there is a T2-hyperintense ring lesion with marked edema (A), and the central portion is strongly hyperintense on DWI (B).
© 2009 by Taylor & Francis Group, LLC
A
B
FIGURE D3.3.48 Herpetic encephalomyelitis. Axial DWI (A) demonstrates hyperintensity in the insulae, visible also on the coronal T2-weighted MR image (B) where hyperintensity is also seen in the right cingulum.
destructive, imaging could have the potential of being superior to autopsy concerning the postmortem detection of malformations. For example, chiari malformation I and II and their complications such as hydrocephalus are well displayed at both CT and MRI. Agenesia of brain structures (e.g., the corpus callosum, vermis), heterotopia of the gray-matter substance and cystic malformations are in most cases detected without problems.
D3.3.8.5 Degenerative Alterations, “White Matter Lesions” D3.3.8.5.1 The Aging Brain The evaluation of myelinization in infants and young adults is complex and requires profound pediatric neuroradiological experience. Although histological sampling actually gives more exact information about the grading of myelinization than imaging, MRI offers the possibility to observe the stage of myelinization by displaying characteristic signal intensities in T1, T2, and inversion-recovery sequences. The aged brain shows typical alterations that are in most cases not associated with functional impairment. In the postmortem examination of forensic cases, these changes are regularly seen and must be distinguished from case-relevant findings. Generally, in elder persons the white-matter substance decreases, and the liquor spaces become larger. This is visible in both CT and MRI. Band-like periventricular hyperintensities are often seen in T2-weighted MRI sequences. The central white-matter substance often shows focal T2-hyperintensities (hypointense in T1, and hypodense at CT imaging) that might cause differential diagnostic problems, as similar white-matter lesions are also observed in potentially symptomatic cerebrovascular diseases such as microangiopathy and lacunar infarction, arteriosclerotic encephalopathy, and chronic hypertension. The perivascular Robin-Virchov spaces are widened in elder patients, which can be seen as T2-hyperintense spots or band-like structures predominantly in the basal ganglia region.
298
The Virtopsy Approach
D3.3.8.5.2 Atrophy and Dementia
A
B
As in the normal aging brain, several forms of dementia are also accompanied by atrophy of the brain structures, which is observed at imaging. Without having information about the clinical status of the person before death, the differential diagnosis between normal aging and pathologic processes will be almost impossible. According to clinical experience [8,40], Mb. Alzheimer atrophy usually includes both hippocampi and the temporal lobes, whereas in Pick’s disease the frontal lobes and frontotemporal regions are the predominantly affected. D3.3.8.5.3 Multiple Sclerosis “Normal” white-matter lesions without impairment of function in the aged brain were described already (see section D3.3.8.5.1). The most common form of pathologic demyelinization besides vascular changes is multiple sclerosis (MS), which can also be observed in forensic autopsies. Although to date no postmortem data exist, clinical MRI shows typical findings in most MS cases (Figure D3.3.49). These are mostly multiple roundish and asymmetric spots that are situated in the periventricular white-matter substance. They appear isointense to hypointense in T1 and hyperintense in T2-weighted and proton-density-weighted sequences. The corpus callosum, internal capsule, and the brain stem are also often affected. In case of neuritis of the opticus nerve, surrounding edema (T2 hyperintense) might be present. Contrast-enhanced imaging would allow further characterization of the lesions and the activity of the illness; however, in postmortem situations this is currently not available.
(C)
FIGURE D3.3.49 Multiple sclerosis is characterized at imaging by typical multiple radiating hyperintensities in the paracallosal white matter at DWI (A), T2 (B), and FLAIR attenuated MRI (C, sagittal; D, coronal).
central pons region is affected (“central pontine myelinolysis”), but extrapontine localizations are also observed in a number of cases (e.g., basal ganglia, periventricular region). Native clinical CT reveals hypodense areas in the affected regions, and MRI shows slightly hypointense regions in T1 that appear hyperintense in the T2-weighted sequences (Figure D3.3.50). In chronic osmotic myelinolysis (after months) CT might display the findings as hyperdense, while T2 hyperintensity in the MRI images decreases. In clinical imaging, DWI can help with the diagnosis as the findings appear hyperintense and the apparent diffusion coefficient ADC is reduced. In contrast to ischemic infarction, the peripheral pontine fibers are not affected in osmotic myelinolysis. As postmortem MRI or CT examinations have not been provided to date in osmotic myelinolysis cases, future forensic radiological research will have to clarify the postmortem imaging characteristics of this finding.
D3.3.8.5.4 Osmotic Myelinolysis, Central Pontine Myelinolysis “Osmotic stress” (i.e., fluctuations of the osmotic gradient) can cause osmotic myelinolysis, which may be detected radiologically. In forensic cases, the main reason for osmotic stress that causes myelinolysis is chronic alcohol abuse. Usually the
A
(D)
B
FIGURE D3.3.50 Osmotic pontine myelinolysis in a living person. The coronal FLAIR (A) and axial T2-weighted MRI (B) show hyperintense signal in the pons.
© 2009 by Taylor & Francis Group, LLC
Incident-Specific Cases
D3.3.9 TOXICOLOGICAL ASPECTS D3.3.9.1 Changes in Chronic Alcohol Abuse Toxic degenerative demyelinization (osmotic myelinolysis, central pontine myelinolysis) in chronic alcohol abuse is described in Section D3.3.8.5.4. Besides secondary alterations due to falls in chronic alcohol abundance, the brain might show signs of atrophy in the hemispheres as well as atrophy of the superior vermis cerebelli, and the ventricles are often enlarged at both CT and MRI. According to the clinical literature [8,40], unspecific T2-hyperintense white-matter lesions can be present in the periventricular regions and the corpus callosum. Wernicke encephalopathy is characterized by T2 hyperintensity of the medial thalamus, the midbrain, and the corpora mamillaria and around the third ventricle. DWI shows restricted diffusion, which appears as hyperintensity in the affected regions. Some new publications point toward a potential of MR spectroscopy to allow the determination of brain alcohol levels [55–57]. However, research concerning the application of MR spectroscopy for the analysis of the grade of alcoholization is still far limited in the living, and postmortem spectroscopic examinations focused on this issue have not yet been performed. D3.3.9.2 Methanol According to the radiological literature [58,59], the bilateral hemorrhagic necrosis of the putamen that is caused by methanol intoxication can be diagnosed at both CT and MRI. Furthermore, hemorrhagic necrosis appearing as white-matter lesions might be found in the subcortical areas and in the visual pathway regions. D3.3.9.3 Drug Abuse Some specific alterations in drug abundance have been observed in recent clinical studies and are thought to occur due to ischemic damage. In cocaine and opiate users, T2hyperintense white-matter lesions were found predominantly in the frontal white matter, while periventricular white-matter lesions were not significantly increased in comparison to healthy persons [60]. In methamphetamine users, similar changes were found [61].
299
cortex and subcortical structures, and the basal ganglia appear hypodense. Both findings are present in almost all postmortem examinations. According to clinical experience, small hemorrhages can be present, and signs of atrophy may be seen in chronic cases. At MRI, the brain tissue including the basal ganglia, the thalamus, and further regions such as the brain stem appears globally hyperintense in the T2-weighted sequences. In acute ischemia, DWI might give sensitive clues for the pathologic process by displaying restricted diffusion, which is also seen as hyperintensity. MRI is in clinical radiology generally superior to CT concerning hypoxic encephalopathy [40].
D3.3.11 EDEMA AND INCREASED BRAIN PRESSURE D3.3.11.1 Edema Edema accompanies many traumatic and nontraumatic findings and is frequently found in forensic autopsy cases. It can be generalized or localized in specific areas. Imaging, especially MRI, depicts edema as follows in clinical as well as in postmortem examinations: In CT, the edematous brain parenchyma appears hypodense, and the differentiation between gray and white matter gets lost. The ventricles and the sulci may be narrowed. Edema due to disturbance of the bloodbrain barrier is more pronounced in the white matter while cytotoxic edema predominantly affects the gray substance. MRI reveals edema as T1-hypointense and T2-hyperintense regions. DWI could become of specific forensic interest since it allows a differentiation between vasogenic and cytotoxic edema based on alterations in the diffusion process [63]: In vasogenic edema, the extracellular water content is increased, causing an increase of the apparent diffusion coefficient ADC. Cytotoxic edema, in contrast, is characterized by a reduced ADC due to predominantly intracellular swelling. The diagnostic accuracy of DWI in postmortem settings, however, has to be clarified in future studies, as the “normal” postmortem ADC decrease that is observed in all postmortem DWI examinations must be taken into account (unpublished Virtopsy data) (Figure D3.3.51). A
B
D3.3.9.4 Carbon Monoxide Intoxication Acute carbon monoxide poisoning leads to characteristic changes that are better visible at MRI than CT. Nonhemorrhagic necrosis of the globus pallidum is known as a characteristic sign of this type of intoxication besides abnormal signal intensities in the other basal ganglia and sometimes also the cortex and hippocampus region [62]. Postmortem imaging of carbon monoxide intoxication has not been performed yet.
D3.3.10 HYPOXIC ENCEPHALOPATHY Following anoxia, typical cerebral alterations can be seen at imaging [8,40]. CT shows reduced differentiation between
© 2009 by Taylor & Francis Group, LLC
FIGURE D3.3.51 Postmortem (A) and vital (B) diffusion-weighted MR image. The ADC values are generally reduced in the postmortem cases, which is reflected by a general hyperintensity in the DW images (white brain).
300
A
The Virtopsy Approach
B
FIGURE D3.3.52 Increased brain pressure with swelling of the cerebellar tonsils (A, arrow) and herniation of the tonsils into the spinal canal (B, arrows, T2-weighted MR image).
D3.3.11.2 Brain Herniation Well known from forensic autopsy, the signs of increased pressure can also be detected postmortem using the imaging methods. It is, however, important to perform and view the scans in diverse planes to also detect the more discrete alterations. Midline shifting and displacement of the ventricles and other liquor spaces can be diagnosed without problems at both CT and MRI (Figure D3.3.23, Figure D3.3.27, and Figure D3.3.29), and imaging is superior to autopsy concerning the diagnosis of liquor congestion. Swelling of the gyri and narrowing of the sulci can also be observed, as well as compression of the liquor spaces and cisterns. Furthermore, the radiological methods allow the diagnosis of displacement of the cingulate gyrus under the falx in subfalcine herniation. In transtentorial herniation, displacement of the cerebellar tonsils, the brain stem, and the vermis can be seen (Figure D3.3.27 and D3.3.52), and the parahippocampal gyrus can be shifted (Figure D3.3.27). In severe herniation parts of the cerebellum and other brain structures might also be displaced. According to the Virtopsy Project, the typical petechial bleedings that can be present in the dislocated and compressed brain structures with few exceptions (Figure D3.3.53) currently remain undetected at CT and MRI in contrast to autopsy. On the other hand, especially MRI allows an early and reliable diagnosis of brain infarction as sometimes seen in severely increased brain pressure due to the compression of blood vessels. Using DWI, restricted perfusion and localized or global hypoxia can be detected, which can only be found by means of histology at autopsy.
D3.3.12 BRAIN DEATH The clinical need of determining brain death in patients who are under intensive care is not relevant in postmortem forensic situations where secure signs of death are always present (livores or rigor mortis, or signs of decomposition). The forensic interest is more focused on characteristic “physiological”
© 2009 by Taylor & Francis Group, LLC
A
B
FIGURE D3.3.53 The same case as in Figure D3.3.23, Figure D3.3.32, and Figure D3.3.34. Axial CT (B) reveals acute subdural hematoma on the right side. Representing classical signs of increased brain pressure, midline shifting to the left as well as a compression of the lateral ventricles were seen at CT (Figure D3.3.23). As a further sign of increased pressure that is known from the autopsy, macroscopic hemorrhage into the gyrus cinguli (A, arrow) is present in this case, which can also be seen on the CT slice (B, arrow).
postmortem alterations and their differentiation from preterminally acquired traumatic or pathologic changes. Based on the Virtopsy Project and other postmortem neuroimaging studies [27,28], postmortem brains generally show some signs of diffuse edema (e.g., reduced gray–white differentiation) and are characterized by hypertensity in the DWI sequences according to a general reduction of the ADC (unpublished Virtopsy data) (Figure D3.3.51). Future studies have to define these postmortem changes and their possible interpretation concerning forensic and other issues. Based on data from the Virtopsy Project, up to at least three days after exitus imaging, however, gives reliable results and an imaging quality at both CT and MRI that might even be superior to the quality obtained in living persons, as movement- and flow-induced artifacts are not present in the dead. D3.3.12.1 “Respirator Brain” Some characteristic intracerebral alterations can be observed in patients who suffered brain death and were under further intensive care for several hours or days. Besides petechial hemorrhage and typical small gray-to-yellow softening lesions, which can be found all over the brain, another known characteristic is softening and liquefying of the brain parenchyma. Radiological experience with imaging of these alterations is, to date, rare. In “respirator brain” cases, common signs of brain death are as previously described, as well as some characteristics of hemorrhage when the extent of petechial hemorrhage is sufficient for being displayed in CT or MRI. According to the Virtopsy Studies, the size of a single lesion more than 3 mm or an accumulation of multiple smaller bleedings situated near to each other allows detection at imaging. If the respirator brain causes typical alterations in DWI is not known to date and will have to be focused in future studies.
Incident-Specific Cases
A
301
D3.3.14 CLINICAL FORENSIC NEUROIMAGING
B
FIGURE D3.3.54 Same case as in Figure D3.3.14. The axial CT (B) shows the right-sided bone defects as well as the extensive decomposition of the soft tissues with liquefaction of the brain and massive gas inclusions in the soft tissues (A, autopsy correlation).
D3.3.13 DECOMPOSITION Decomposition processes cause characteristic time- and temperature-dependent alterations of the intracranial tissues such as liquefying and gas intrusion. Some primary experiences with imaging of early and late stages of decomposition have been made in the Virtopsy Project (see Chapter D2.1). Interestingly and important for forensic practice, even in late stages of decomposition relevant findings can be present at imaging (Figure D3.3.54) and can even be better displayed and diagnosed than at autopsy. In a case of a person falling from a staircase and who had been found in a severely decomposed status, frontal hemorrhage was seen at MRI and could be localized far more exactly than at autopsy where the brain was liquefied and showed red-brown and gray discoloration (Figure D3.3.55). Although experience with forensic imaging in these cases is limited, the potential of CT and MRI is obvious. A
B
FIGURE D3.3.55 Frontal intracerebral traumatic hemorrhage in a far decomposed body following a fall from stairs in a far decomposed case. The hemorrhage is displayed well at MRI (B), and, in contrast to autopsy where the examination technique causes alteration of the finding with blood smearing (A), the extent of the finding can be estimated perfectly using the imaging data (6).
© 2009 by Taylor & Francis Group, LLC
Forensic pathologists are regularly confronted with radiological findings in cases who survived neurotrauma and who were treated in a hospital. In these cases, a (neuro)radiologist is usually involved in the clinical process to provide the radiological diagnoses. A close collaboration between the radiologist and the forensic examiner is indispensable to get an optimized basis also for the forensic expertise. Forensic education of radiologists, as well as some knowledge of the forensic pathologist about basic imaging principles, will improve the value of such collaboration in the future, especially when it comes to addressing specific issues like forensic reconstruction or the evaluation of the ability to act following injury. It is likely that radiological examinations will in the future be increasingly requested for forensic purposes only, even when there is no clinical indication. However, the ethical and juridical basis for such nonclinical examinations has to be developed and provided in advance (see Section D3.11). A specialized application of MRI that has some potential to become relevant also in clinical forensic neuroimaging is functional MRI (fMRI). Predominantly in the field of forensic psychiatry there are some possible future applications of this method. First, studies concerning behavioral issues and the evaluation of fMRI as a “lie detector” are currently conducted [64,65]. However, fMRI in forensic psychiatry contains a dangerous potential, and ethical issues are of great importance concerning such applications. The inevitable future discussion and research will determine the forensic value of the method.
D3.3.15 FUTURE ASPECTS CT and MRI are today’s gold standard in the clinical neuroradiological examination of living patients. Although well established in clinical use, the application of these imaging methods for forensic purposes is relatively new, and to date prospective evaluation studies for the postmortem CT and MR examination of the cranium are still rare—or even missing. However, the research that has been conducted to date points toward a promising potential for radiological methods: Besides the general advantages of noninvasive examination (Chapter A), postmortem neuroimaging provides an excellent overview of the findings that can be examined in all planes and in their relation to each other. Regions difficult to investigate at autopsy, such as the neck region or the orbita, become easily accessible by imaging. Forensic reconstruction based on CT and MRI has turned out to be superior to autopsy (Chapters B2, B3, C3, and D3), which is mainly due to the excellent possibilities of handling and visualizing the digital data. The problem of imaging the skin and peripheral cranial soft tissues might be improved by means of surface scanning that could be used in combination with the radiological methods. In comparison with clinical imaging, there is, however, a lack of examining blood vessels in postmortem scans. Due to the absent blood flow, the general imaging quality is often better than in living patients, as flow artifacts are not present; on the other hand, angiographic methods would be required
302
The Virtopsy Approach
A
B
C
D
FIGURE D3.3.56 New diagnostic possibilities using DTI: Fiber tractography can be reconstructed and displayed (A,B) from the anisotropy maps (C,D), offering new and exciting insights into neurotraumatic processes.
for sufficiently visualizing various forms of injury associated with the vessels. The development of new technologies in this field is addressed in current studies [66,67], as it is with the evaluation of postmortem image-guided biopsy tools for the sampling of tissue specimens for histological expertise [68] (see Chapters D5 and D6). The rapid technical developments in clinical radiology will also support the forensic progress (Figure D3.3.56), and the application of new technologies such as dual-source CT and high-Tesla MRI will become possible in forensic settings. Image resolution will be improved by these techniques, and special applications such as whole-body MRI will become available also for routine forensic use. The implementation of forensic neuroimaging based on clinical radiological developments has the potential to amend the basis of forensic expertise in the future and therefore to generally improve legal certainty.
D3.3.16 ACKNOWLEDGMENTS The authors thank Professor Richard Dirnhofer and the Virtopsy Team from the Institute of Forensic Medicine in Bern, Switzerland. Many thanks also to the members of the clinical Neuroimaging Department and the Department of MR Spectroscopy and Methodology of the University Hospital Bern. Cordial thanks go to Dr. Karl-Olof Lövblad for providing the clinical CT and MRI images.
D3.3.17 REFERENCES 1. Langlois JA, Rutland-Brown W, and Thomas KE. Traumatic Brain Injury in the United States: Emergency Department Visits, Hospitalizations, and Deaths. Centers for Disease Control and Prevention, National Center for Injury Prevention and Control, Atlanta, 2004.
© 2009 by Taylor & Francis Group, LLC
2. Adekoya N, Thurman DJ, White DD, and Webb KW. Surveillance for traumatic brain injury deaths—United States, 1989–1998. MMWR Surveill Summ 2002; 51(10):1–14. 3. Zygun DA, Laupland KB, Hader WJ, Kortbeek JB, Findlay C, Doig CJ, et al. Severe traumatic brain injury in a large Canadian health region. Can J Neurol Sci 2005; 32(1):87–92. 4. Masson F, Thicoipe M, Aye P, Mokni T, Senjean P, Schmitt V, et al. Aquitaine Group for Severe Brain Injuries Study. Epidemiology of severe brain injuries: a prospective population-based study. J Trauma 2001; 51(3):481–89. 5. Sundstrom T, Sollid S, and Wester K. [Deaths from traumatic brain injury in the Nordic countries, 1987–2000]. Tidsskr Nor Laegeforen 2005; 125(10):1310–12. 6. Steudel WI, Cortbus F, and Schwerdtfeger K. Epidemiology and prevention of fatal head injuries in Germany—trends and the impact of the reunification. Acta Neurochir (Wien) 2005; 147(3):231–42. 7. Randall BB, Fierro MF, and Froede RC. Practice guideline for forensic pathology. Arch Pathol Lab Med 1998; 122:1056–64. 8. Osborn AG, ed. Diagnostic Neuroradiology. Mosby, St. Louis, 1994. 9. Gentry LR. Imaging of closed head injury. Radiology 1994; 191:1–17. 10. Boyko OB, Alston SR, Fuller GN, Hulette CM, Johnson GA, and Burger PC. Utility of postmortem magnetic resonance imaging in clinical neuropathology. Arch Pathol Lab Med 1994; 118(3):219–25. 11. Thali MJ, Yen K, Schweitzer W, Vock P, Boesch C, Ozdoba C, et al. Virtopsy, a new imaging horizon in forensic pathology: virtual autopsy by postmortem multislice computed tomography (MSCT) and magnetic resonance imaging (MRI)—a feasibility study. J Forensic Sci 2003; 48(2):386–403. 12. Patriquin L, Kassarjian A, Barish M, Casserley L, O’Brien M, Andry C, et al. Postmortem whole-body magnetic resonance imaging as an adjunct to autopsy: preliminary clinical experience. J Magn Reson Imaging 2001; 13(2):277–87.
Incident-Specific Cases
13. Ros PR, Li KC, Vo P, Baer H, and Staab EV. Preautopsy magnetic resonance imaging: initial experience. Magn Reson Imaging 1990; 8(3):303–08. 14. Hart BL, Dudley MH, and Zumwalt RE. Postmortem cranial MRI and autopsy correlation in suspected child abuse. Am J Forensic Med Pathol 1996; 17(3):217–24. 15. Paperno S, Riepert T, Krug B, Rothschild MA, Schultes A, Staak M, et al. [Value of postmortem computed tomography in comparison to autopsy]. Rofo 2005; 177(1):130–36. 16. Donchin Y, Rivkind AI, Bar-Ziv J, Hiss J, Almog J, and Drescher M. Utility of postmortem computed tomography in trauma victims. J Trauma 1994; 37(4):552–55. 17. Thali MJ, Yen K, Vock P, Ozdoba C, Kneubuehl BP, Sonnenschein M, et al. Image-guided virtual autopsy findings of gunshot victims performed with multi-slice computed tomography and magnetic resonance imaging and subsequent correlation between radiology and autopsy findings. Forensic Sci Int 2003; 138(1–3):8–16. 18. Schumacher M, Oehmichen M, Konig HG, and Einighammer H. [Intravital and postmortal CT examinations in cerebral gunshot injuries]. ROFO Fortschr Geb Rontgenstr Nuklearmed 1983; 139(1):58–62. 19. Schumacher M, Oehmichen M, Konig HG, Einighammer H, and Bien S. [Computed tomographic studies on wound ballistics of cranial gunshot injuries]. Beitr Gerichtl Med 1985; 43:95–101. 20. Yen K, Sonnenschein M, Thali MJ, Ozdoba C, Weis J, Zwygart K, et al. Postmortem multislice computed tomography and magnetic resonance imaging of odontoid fractures, atlantoaxial distractions and ascending medullary edema. Int J Legal Med 2005; 119(3):129–36. 21. Aghayev E, Yen K, Sonnenschein M, Ozdoba C, Thali M, Jackowski C, et al. Virtopsy post-mortem multi-slice computed tomography (MSCT) and magnetic resonance imaging (MRI) demonstrating descending tonsillar herniation: comparison to clinical studies. Neuroradiology 2004; 46(7):559–64. 22. Ozdoba C, Weis J, Plattner T, Dirnhofer R, and Yen K. Lethal scuba diving accident with massive air embolism in cerebral and spinal arteries. Am J Neuroradiol 2004: 47(6):411–16. 23. Stein KM, Bahner ML, Merkel J, Ain S, and Mattern R. Detection of gunshot residues in routine CTs. Int J Legal Med 2000; 114(1–2):15–18. 24. Oliver J, Lyons TJ, and Harle R. The role of computed tomography in the diagnosis of arterial gas embolism in fatal diving accidents in Tasmania. Australas Radiol 1999; 43(1):37–40. 25. Iwama T, Andoh H, Murase S, Miwa Y, and Ohkuma A. Diffuse cerebral air embolism following trauma: striking postmortem CT findings. Neuroradiology 1994; 36(1):33–34. 26. Iwase H, Yamada Y, Ootani S, Sasaki Y, Nagao M, Iwadate K, et al. Evidence for an antemortem injury of a burned head dissected from a burned body. Forensic Sci Int 1998; 94(1–2):9–14. 27. Englund E, Sjobeck M, Brockstedt S, Latt J, and Larsson EM. Diffusion tensor MRI post mortem demonstrated cerebral white matter pathology. J Neurol 2004; 251(3):350–52. 28. Lovblad KO, Bassetti C, and Basssetti C. Diffusionweighted magnetic resonance imaging in brain death. Stroke 2000; 31(2):539–42. 29. Yen K, Lovblad KO, Scheurer E, Ozdoba C, Thali M, Aghayev E, et al. Post-mortem forensic neuroimaging: correlation of MSCT and MRI findings with autopsy results. Forensic Sci Int 2007; 173(1):21–35.
© 2009 by Taylor & Francis Group, LLC
303
30. Harris LS. Postmortem magnetic resonance images of the injured brain: effective evidence in the courtroom. Forensic Sci Int 1991; 50(2):179–85. 31. Fernando MS, O’Brien JT, Perry RH, English P, Forster G, McMeekin W, et al. Comparison of the pathology of cerebral white matter with post-mortem magnetic resonance imaging (MRI) in the elderly brain. Neuropathol Appl Neurobiol 2004; 30(4):385–95. 32. Blamire AM, Rowe JG, Styles P, and McDonald B. Optimising imaging parameters for post mortem MR imaging of the human brain. Acta Radiol 1999; 40(6):593–97. 33. Messori A and Salvolini U. Postmortem MRI as a useful tool for investigation of cerebral microbleeds. Stroke 2003; 34(2):376–77. 34. Shen WC, Shieh TT, Shih TP, Chang CY, Su MC, Lee SK, et al. [MRI of postmortem brains]. Gaoxiong Yi Xue Ke Xue Za Zhi 1993; 9(12):690–97. 35. Whitby EH, Paley MN, Cohen M, and Griffiths PD. Postmortem MR imaging of the fetus: An adjunct or a replacement for conventional autopsy? Semin Fetal Neonatal Med 2005; 10:420–26. 36. Woodward PJ, Sohaey R, Harris DP, Jackson GM, Klatt EC, Alexander AL, et al. Postmortem fetal MR imaging: comparison with findings at autopsy. Am J Roentgenol 1997; 168(1):41–46. 37. Pape KE, Bennett-Britton S, Szymonowicz W, Martin DJ, Fitz CR, and Becker L. Diagnostic accuracy of neonatal brain imaging: a postmortem correlation of computed tomography and ultrasound scans. J Pediatr 1983; 102(2):275–80. 38. Ludwig B, Becker K, Rutter G, Bohl J, and Brand M. Dyke Postmortem CT and Award: autopsy in perinatal intracranial hemorrhage. Am J Neuroradiol 1983; 4(1):27–36. 39. Griffiths PD, Variend D, Evans M, Jones A, Wilkinson ID, Paley MN, et al. Postmortem MR imaging of the fetal and stillborn central nervous system. Am J Neuroradiol 2003; 24(1):22–27. 40. Brookes JA, Hall-Craggs MA, Sams VR, and Lees WR. Non-invasive perinatal necropsy by magnetic resonance imaging. Lancet 1996; 348(9035):1139–41. 41. Parizel PM, Tanghe H, Hofman PAM, and Puskas Z. Kernspinto-mographie des Gehirns. In: Reimer P, Parizel PM, and Stichnoth FA, eds. Klinische MR-Bildgebung: eine praktische Anleitung. Springer, Berlin, 2000, 69–137. 42. Parizel PM, Makkat S, Jorens PG, Ozsarlak O, Cras P, Van Goethem JW, et al. Brainstem hemorrhage in descending transtentorial herniation (Duret hemorrhage). Intensive Care Med 2002; 28(1):85–88. 43. Yen K, Weis J, Kreis R, Aghayev E, Jackowski C, Thali M, et al. Line-scan diffusion tensor imaging of the posttraumatic brain stem: changes with neuropathologic correlation. Am J Neuroradiol 2006 27(1):70–73. 44. Brinkmann B and Madea B. Handbuch Gerichtliche Medizin Bd. 1. Springer, Berlin, 2003. 45. Unterharnscheidt F. Gedeckte Schäden des Gehirns. In: Unterharn-scheidt F, ed. Pathologie des Nervensystems VI—Trauma-tologie von Hirn und Rückenmark. Traumatische Schäden des Gehirns (forensische Pathologie). Springer, Berlin, 1993. 46. Kleinman PK. Diagnostic Imaging of Child Abuse. Elsevier, Oxford, 1998. 47. Jackowski C, Thali M, Sonnenschein M, Aghayev E, Yen K, Dirnhofer R, et al. Visualization and quantification of air embolism structure by processing postmortem MSCT data. J Forensic Sci 2004; 49(6):1339–42.
304
48. Stoeger A, Daniaux M, Felber S, Stockhammer G, Aichner F, and Nedden D. MRI findings in cerebral fat embolism. Eur Radiol. 1998; 8(9):1590–93. 49. Sakamoto T, Sawada Y, Yukioka T, Yoshioka T, Sugimoto T, and Taneda M. Computed tomography for diagnosis and assessment of cerebral fat embolism. Neuroradiology 1983; 24(5):283–85. 50. Marshall GB, Heale VR, Herx L, Abdeen A, Arkonjic L, Powell J, et al. Magnetic resonance diffusion weighted imaging in cerebral fat embolism. Can J Neurol Sci 2004; 31(3): 417–21. 51. Butteriss DJ, Mahad D, Soh C, Walls T, Weir D, and Birchall D. Reversible cytotoxic cerebral edema in cerebral fat embolism. Am J Neuroradiol 2006; 27(3):620–23. 52. Gorrino O, Oleaga L, de la Fuente R, Garcia Bolado A, and Grande D. Intracranial artefact in magnetic resonance caused by embolization of microscopic metallic fragment. Neurologia 2004; 19(10):766–68. 53. Anda T, Suyama K, Kawana T, and Mori K. Shotgun pellet embolus in the cerebral circulation via the internal carotid artery in the neck; a case report. No Shinkei Geka 1992; 20(4):457–61. 54. Muir KW, Buchan A, von Kummer R, Rother J, and Baron JC. Imaging of acute stroke. Lancet Neurol 2006; 5(9):755–68. 55. Lovblad KO. Diffusion-weighted MRI: back to the future. Stroke 2002; 33(9):2204–05. 56. Fein G and Meyerhoff DJ. Ethanol in human brain by magnetic resonance spectroscopy: correlation with blood and breath levels, relaxation, and magnetization transfer. Alcohol Clin Exp Res 2000; 24(8):1227–35. 57. Sammi MK, Pan JW, Telang FW, Schuhlein D, Molina PE, Volkow ND, et al. Measurements of human brain ethanol T(2) by spectroscopic imaging at 4 T. Magn Reson Med 2000; 44(1):35–40. 58. Hetherington HP, Telang F, Pan JW, Sammi M, Schuhlein D, Molina P, et al. Spectroscopic imaging of the uptake kinetics of human brain ethanol. Magn Reson Med 1999; 42(6):1019–26. 59. Sefidbakht S, Rasekhi AR, Kamali K, Borhani Haghighi A, Salooti A, Meshksar A, et al. Methanol poisoning: acute MR and CT findings in nine patients. Neuroradiology 2007; 49(5):427–35. 60. Blanco M, Casado R, Vazquez F, and Pumar JM. CT and MR imaging findings in methanol intoxication. Am J Neuroradiol 2006; 27(2):452–54. 61. Lyoo IK, Streeter CC, Ahn KH, Lee HK, Pollack MH, Silveri MM, et al. White matter hyperintensitis in subjects with cocaine and opiate dependence and healthy comparison subjects. Psychiatry Res 2004; 131(2):135–45. 62. Bae SC, Lyoo IK, Sung YH, Yoo J, Chung A, Yoon SJ, et al. Increased white matter hyperintensities in male methamphetamine users. Drug Alcohol Depend 2006; 81(1):83–88. 63. O`Donnell P, Buxton PJ, Pitkin A, and Jarvis LJ. The magnetic resonance imaging appearances of the brain in acute carbon monoxyde poisoning. Clin Radiol 2000; 55(4):273–80. 64. Schwarcz A, Ursprung Z, Berente Z, Bogner P, Kotek G, Meric P, et al. In vivo brain edema classification: new insight offered by large b-value diffusion-weighted MR imaging. J Magn Reson Imaging 2007; 25(1):26–31. 65. Mohamed FB, Faro SH, Gordon NJ, Platek SM, Ahmad H, and Williams JM. Brain mapping of deception and truth telling about an ecologically valid situation: functional MR imaging and polygraph investigation—initial experience. Radiology 2006; 238(2):679–88.
© 2009 by Taylor & Francis Group, LLC
The Virtopsy Approach
66. Langleben DD, Loughead JW, Bilker WB, Ruparel K, Childress AR, Busch SI, et al. Telling truth from lie in individual subjects with fast event-related fMRI. Hum Brain Mapp 2005; 26(4):262–72. 67. Grabherr S, Djonov V, Friess A, Thali MJ, Ranner G, Vock P, et al. Postmortem angiography after vascular perfusion with diesel oil and a lipophilic contrast agent. Am J Roentgenol 2006; 187(5):W515–23. 68. Jackowski C, Sonnenschein M, Thali M, Aghayev E, v. Allmen G, Yen K, et al. Virtopsy: postmortem minimally invasive angiography using cross section techniques— implementation and preliminary results. J Forensic Med 2004; 50(5):1175–86. 69. Aghayev E, Thali MJ, Sonnenschein M, Jackowski C, Dirnhofer R, and Vock P. Post-mortem tissue sampling using computed tomography guidance. Forensic Sci Int. 2007; 166(2–3):199–203.
D3.4 SHARP TRAUMA Stephan A. Bolliger and Michael J. Thali
D3.4.1 INTRODUCTION Sharp trauma plays a large role in everyday forensic practice, be it in a homicidal, suicidal, or accidental setting. The main common denominator of the injury-inflicting objects in sharp trauma is that they can pierce and slice the human body and thus inflict internal damage. Death due to harming the body’s integrity is manifold; exsanguination due to injury of blood vessels is the most frequently encountered form. However, death due to, for example, air embolism and hemopericardium is also encountered regularly. In the assessment of sharp trauma, issues such as the number and location of the injuries, the wound morphology, the wound channel, the injuries inflicted to the soft tissues and the skeleton, and the cause of death should always be addressed in order to evaluate possible third-party involvement and to undertake an incident reconstruction. Depending on the energy applied, sharp violence may lead to the severing of limbs, as seen in hacks with machetes or axes. Although injuries due to blows with axes and machetes are rare in the Western world, knife wounds and wounds due to other sharp objects such as shattered glass in traffic accidents are frequently seen. It is only natural that one utilizes a well-known object of everyday life to dispose of an enemy or competitor. Therefore, it is not surprising that knives have retained their popularity throughout the millennia for settling disputes. It may be indeed stated that of all the forms of violence due to a third party, sharp trauma plays a cardinal role in mankind’s past and present. Sharp trauma can be divided into three categories: stab wounds, cuts, and chops or hacks.
D3.4.2 STAB WOUNDS These injuries are inflicted by pointed, sharp instruments, generally in homicidal situations. A stab injury is characterized
Incident-Specific Cases
by a wound that is deeper than it is wide. The wound margins are sharp, generally straight, not undermined, and devoid of bruising. Depending on the instrument or weapon, one end of the wound may present an abrasion resulting from the back of the blade. If the blade is plunged into the victim deeply enough, the hilt of the knife may give rise to an abrasion. These abrasions are enormously important, as they can give clues as to the type of knife involved. Stab wounds should be clearly separated from wounds due to impaling, which also give rise to deep and small wounds but are inevitably accompanied by an abrasion and are rarely homicidal but rather accidental. Objects that inflict stab injuries are, as the name implies, sharp and encompass not only the obvious such as knives, daggers, and scissors but also glass shards, screwdrivers, cutlery, and even writing utensils, to name just a few. The examination and interpretation of stab wound morphology is of great importance in forensic pathology. The documentation of such wounds is generally performed by photography. However, this technique reduces a three-dimensional injury to a two-dimensional image, thus losing the 3D information. 3D true-color photogrammetry-based surface scanning, on the other hand, can document the lesion in a three-dimensional fashion for future reference (see Chapter B1, “External Body Documentation”). However, multislice computed tomography (MSCT) is also suitable for the documentation of such injuries, although it is of inferior quality compared with the aforementioned techniques with regard to precisely depicting wound morphology. It is, however, very rapid, gives a general impression of the size and location on the body, and depicts it in a three-dimensional manner (Figure D3.4.1, Figure D3.4.2, and Figure D3.4.3). As wound morphology belongs to the realm of external examination, it is therefore not discussed further in this book. Apart from external appearance, which is readily recognized by any person accustomed to forensic pathology and can give an essentially critical clue as to the shape and type of inflicting instrument, stab wounds pose other questions to the forensic examiner. Apart from the cardinal questions of third-party involvement—a topic dealt with in detail in innumerous forensic pathology textbooks—these may be of reconstructive nature. Such reconstructive questions are, for example, the stab direction, stab depth, stab energy, weapon handling, and weapon type. Depending on the direction of a stab wound, different reconstructive patterns may arise and can prove a suspect or witness to be credible or not. We have often been confronted with a suspect’s version of the incident being committed in self defense, whereas the stab wound direction literally pointed to the opposite. Rarely, the knife is still embedded in the victim. Obviously, in such cases where the stab direction is clearly given by the weapon, the removal of the knife would only complicate the forensic assessment. A CT scan can depict the embedded knife in situ and, thus, the stab direction with great precision [1]. Alas, the still-embedded knife is a rarity in everyday forensic pathology. Therefore, in traditional forensic
© 2009 by Taylor & Francis Group, LLC
305
FIGURE D3.4.1 Autopsy photograph (top) and 3D MSCT surface reconstruction of a 15-year-old male who committed suicide by ramming a kitchen knife into his chest. The stab wound is clearly visible (arrow).
pathology, such stab wound directions are examined either by inserting a probe or otherwise comparable object into the wound or by painstakingly dissecting layer by layer of the surrounding tissue. The first method is obsolete as it either may harm previously intact structures or may displace traces into the depth of the wound. The second option—layer-bylayer dissection—is certainly superior to the probing of the wound but is excessively time-consuming. Furthermore, intact structures in the immediate vicinity of the wound channel are destroyed, thus complicating the differentiation between perimortal and postmortal injury. Certain additional findings, which may be immensely important in the assessment of the vitality of the stabbing (e.g., whether the victim was dead upon being stabbed) can also be destroyed. Gas embolism, an often encountered and vital sign in stabbing, is almost impossible to detect upon opening the body (see Chapter D2.3, “Vital Reactions and Vital Signs”).
306
FIGURE D3.4.2 Autopsy photo (top) and MSCT 3D surface reconstruction showing a suicidal stab wound (arrow) to the chest of an elderly male.
Postmortem CT is helpful in determining the general stab direction but lacks somewhat in depicting the actual wound channel through soft tissues. The latter is often only seen due to the secondary influx of air or accompanying hemorrhage (Figure D3.4.4). However, postmortem CT is ideal for the detection of osseous or cartilaginous lesions. These structures are more or less frequently affected [2,3] (Figure D3.4.5, Figure D3.4.6, and Figure D3.4.7). Postmortem magnetic resonance imaging (MRI), the method of choice for the depiction of soft tissues and their lesions (Figure D3.4.8 and Figure D3.4.9), can image the wound channel through the soft tissues with great accuracy and therefore can determine in a nondestructive fashion the course of the wound channels. Thus, the previously unharmed structures retain their integrity, which facilitates the assessment of actual stab injuries in comparison with autoptic lesions, which are obviously also due to sharp trauma. Furthermore, a virtual probe can be inserted into the wound without displacing traces. Apart from the fact that such a virtual stab direction can be reproduced time and time again
© 2009 by Taylor & Francis Group, LLC
The Virtopsy Approach
FIGURE D3.4.3 The chest of a female victim of a homicidal stabbing. The closely grouped, predominantly uniform wounds shown on the autopsy photograph (top) and the MSCT surface reconstruction (bottom) initially awoke the assumption of a suicide. However, the findings at the scene ruled out this possibility.
(which a layer-by-layer dissection obviously cannot), the virtual probe adheres closer to the real body’s morphology. The course of stab wounds is, by virtue of their thin, mostly slit-like appearance, difficult to assess during external examination. This aspect, which is easier to judge in cuts and chops (see next section), is often a cardinal question. Even if of sufficient depth, a stab to the chest may miss all immediately vital organs completely. Other, seemingly harmless stab wounds can pose an immediately life-threatening situation. Surprising stab courses occur occasionally. For example, in a case of a homicidal stabbing of a young man by his girlfriend (Figure D3.4.10, Figure D3.4.11, Figure D3.4.12, and Figure D3.4.13), two stab wounds were evident: one to the right side of the chest and the other to the right shoulder. The chest stab was confined to the subcutaneous tissue, while the shoulder stab harmed both the brachial vein and the artery. The cause of death was exsanguination due to the arterial lesion and not—as was presumed at the crime scene—a pneumothorax or internal hemorrhage. Especially in cases of multiple stab wounds, the ultimately lethal stab is of utmost importance. For example, if a victim displays multiple, but
Incident-Specific Cases
307
FIGURE D3.4.4 MSCT sagittal reconstruction of the thorax of the case shown in Figure D3.4.1. Note the stab channel (the boundaries of which are designated by two dashed lines) through the chest into the immediate vicinity of or into the heart. The small image is an autopsy photo of the opened rib cage with a probe demonstrating the wound channel. FIGURE D3.4.6 MSCT 3D bone reconstruction, lateral view. In this image, the osseous lesions depicted in Figure D3.4.5 are seen from a lateral angle, thus facilitating the geometric assessment of the stab region.
FIGURE D3.4.5 MSCT 3D bone reconstruction, front view of the case seen in Figure D3.4.3. In this image, only large lesions are evident, as the computer program automatically smoothes finer structures, thus obliterating hairline fractures. However, a fracture of the third rib (red arrow) and the grooves (yellow arrows) due to the stabs are clearly depicted.
© 2009 by Taylor & Francis Group, LLC
FIGURE D3.4.7 Photograph of the cleaned rib cage depicted in Figure D3.4.5 and Figure D3.4.6. The grooves and fractures seen on the radiological images could be confirmed.
308
The Virtopsy Approach
FIGURE D3.4.8 T2-weighted sagittal MRI of the chest. The left ventricle displays a lesion (yellow arrow). In the pericardial sack and the thoracic cavity, signal-rich layer (red arrows) corresponding to the cell-poor blood components is depicted. The cellular components of the blood are seen in the dependant (in this case, posterior) regions (X).
not deadly, stab wounds to the chest and a lethal stab wound to the back, the presumption of a solely defensive stabbing may be seriously questioned. No human is built equally; however, schematic illustrations do not take this into account, as they merely serve to give a
FIGURE D3.4.9 T2-weighted axial MRI of the chest. In the left thoracic cavity, a signal-rich layer (X) corresponding to the cellpoor blood components and the signal-poor (XX) layer of blood cells are depicted. The inserted autopsy photo demonstrates that the left-sided hemothorax has coagulated.
© 2009 by Taylor & Francis Group, LLC
FIGURE D3.4.10 Autopsy photograph of the right shoulder region of a 30-year-old man who was stabbed twice by his girlfriend during a scuffle and collapsed several minutes later. She claimed that during the fight he lurched toward her and plunged into a knife that she had held in front of her in order to ward him off. The yellow probes indicate the wound channel direction.
general and simplified overview. In postmortem CT reconstructions of stab wound channels, the body is depicted accurately in relation to proportions and height. It plays an imminently important role in assessing the depth of a wound if the thickness of the subcutaneous fatty tissue is known. To puncture an obese person’s abdomen, a far deeper stab wound is necessary than, for instance, in the case of a lean child. The contrary is also true in cases in which the blade is not completely thrust into the victim. Modern technology such as postmortem CT can measure such depths with great precision. However, great caution should be exercised in interpreting the wound depth; as the human skin is elastic and the underlying soft tissues are—as the name implies—soft, the inflicting instrument may be shorter than the actual wound channel. Despite these caveats, the depth of the wound can give rise to a general impression of the incident. Human skin is tough and, as surely every pathologist can confirm, requires a certain amount of energy to pierce. This has been examined by several authors in the past [4–7]. Thus, an extremely deep stab wound does hold clues as to the perpetrator’s emotional situation. The overall impression of the stab depth, although not a “hard” clue, is also important in the reconstruction of a crime scene. Postmortem MSCT is helpful in determining the thickness of the punctured tissues such as fatty tissue and musculature
Incident-Specific Cases
309
FIGURE D3.4.13 MSCT axial reconstruction of the right shoulder. Note the stab wound to the chest (green arrow) and to the arm (yellow arrow). The pleural cavity is intact.
FIGURE D3.4.11 Autopsy photo depicting the injured brachial artery and its corresponding vein. These blood-vessel injuries led to death by exsanguination.
and therefore the overall wound depth (Figure D3.4.14). Furthermore, as previously mentioned, injuries to osseous or cartilaginous structures are detected by postmortem MSCT. According to Banasr et al. [2], injuries to bones and cartilage occur in 53% of cases of sharp trauma. In this study these
FIGURE D3.4.12 MSCT coronal reconstruction of the right shoulder. Note that the wound channel (yellow arrow) of the stab to the chest did not reach the pleural cavity.
FIGURE D3.4.14 MSCT axial reconstruction. Note the large hemothorax (X) and the stab wounds (arrows). The total distance between the chest skin and the posterior pleural lining—in this case the maximal distance of the stab channels—can be measured (dotted line).
© 2009 by Taylor & Francis Group, LLC
310
lesions were encountered predominantly at the rib cage, the laryngeal cartilages, and the spinal column. Such telltale marks help in assessing the case for two reasons. First, it takes a greater amount of energy to penetrate a bone. A careful assessment of the energy needed to inflict the stab is thus possible. Second, the hard structures of the body give a more accurate image of the weapon than soft tissues, which may be distorted due to their elasticity or may have simply rotted away. A groove may give clues as to whether a knife blade’s back or cutting edge struck the bone or cartilage, thus permitting a reconstruction of the knife’s position when stabbing the victim. This in turn is helpful in determining how the assailant stabbed the victim. Another advantage of postmortem MCT is the possibility of reconstructing a case upon which surgery has been undertaken. For instance, in a case of a stabbing with a pocket knife to the head and trunk, the severely injured victim was brought to the hospital for emergency surgery. The surgeons performed an osteoclastic brain surgery in order to evacuate a stab-inflicted hematoma. The victim died three days after the intervention. Unfortunately, the removed skull piece was discarded. The clinical, antemortem MSCT data could deliver the necessary data as to the number of head stabs and their depth (Figure D3.4.15, Figure D3.4.16, Figure D3.4.17, Figure D3.4.18, and Figure D3.4.19). Although the CT technology today does not give a sufficiently accurate image of such grooves, it certainly helps in detecting such lesions prior to autopsy. The seeking of these telltale marks is therefore facilitated at autopsy. However, we believe that the rapid development of the current MSCT
FIGURE D3.4.15 MSCT 3D reconstruction of the clinical, premortem data of a victim of a homicidal stabbing to the head and trunk. Two osseous lesions (red and blue arrows) and a surgical clipping (yellow arrow) are clearly visible.
© 2009 by Taylor & Francis Group, LLC
The Virtopsy Approach
FIGURE D3.4.16 MSCT 3D reconstruction of the postmortem data of the victim seen in Figure D3.4.15. Note the skull defect due to the osteoclastic craniotomy (red arrows) and the surgical stapling of a further wound (yellow arrow).
FIGURE D3.4.17 Postmortem MSCT, coronar image of the head of the victim seen in Figure D3.4.15 and Figure D3.4.16. The stab channel (demarcated by yellow lines) and blood within the ventricular system (asterisk) are clearly visible. Note also the skull defect due to craniotomy (red arrow) with the brain bulging out and a midline shift toward the right (green arrows) due to an increasing and ultimately lethal cerebral edema.
Incident-Specific Cases
311
technology will permit an accurate imaging of osseous lesions in the corpse in the near future. Other, albeit rare, findings in a forensic pathologists routine are foreign bodies. As a knife blade—which is made for cutting, not stabbing—is usually slender and delicate at the tip, this tip may break upon striking a hard object such as bone. Such knife tips can give clues not only as to the type of weapon but also to which knife was involved. As this knowledge may be decisive in solving a case, great care should be taken in detecting these telltale remnants. Although common-plane x-rays, performed daily in most forensic institutes, give a good idea regarding the presence of such objects, postmortem CT serves to depict their localization in a three-dimensional fashion, thus facilitating the localization at autopsy. Furthermore, if the object is large enough (i.e., exceeding a few millimeters), the CT scan may even serve to create a three-dimensional reconstruction of the object, thus permitting a comparison of the presumed knife to the visualized knife tip. In summary, postmortem imaging is a useful tool in evaluating cases in which a victim was stabbed to death. The following points discussed herein are recapitulated briefly as follows: FIGURE D3.4.18 Postmortem MSCT 3D depiction of the head of the victim seen in Figure D3.4.15, Figure D3.4.16, and Figure D3.4.17. The skull cap has been virtually removed in order to show the osteoclastic defect (red arrow) and the stab channel (demarcated by yellow lines).
r MSCT can give an overall impression of the general stab depth. This can give clues as to how much energy was applied. The stab depth should not be confused with the length of a blade, as a remarkable difference is not only possible but is often the rule. r MSCT can depict lesions to the hard tissues such as the skeleton and the cartilage. On these hard structures, conclusions as what type of blade was used are often possible. Furthermore, it is often possible to assess the knife’s position with respect to the localization of the blade back and cutting edge. r MSCT is, in our opinion, the method of choice in detecting foreign bodies in the corpse. Should these belong to the inflicting instrument, such as a broken knife tip, then not only can the type of weapon be judged, but also a unique instrument can be attributed to the stabbing. r MSCT can give clues as to the cause of death (e.g., hemorrhage, gas embolism). r MRI is more suitable for the assessment of softtissue lesions. With this technology, wound channels can be demonstrated with great accuracy, thus giving essential clues as to the stabbing situation. If the wound channel and therefore the stab direction are known, an alleged perpetrator’s version of the incident may be proven or disproved.
D3.4.3 CUTS FIGURE D3.4.19 Clinical MSCT with contrast agent, axial image through the abdomen of the victim seen in Figure D3.4.15, Figure D3.4.16, Figure D3.4.17, and Figure D3.4.18. A stab channel through the abdominal wall (red arrows) is clearly depicted. Note also the cyst in the left kidney (asterisk).
© 2009 by Taylor & Francis Group, LLC
Cuts are created by sharp-edged objects, typically knives. As mentioned already, other sharp-edged objects such as glass shards suffice to create such potentially life-threatening wounds. Although estimation of the depth of a cut wound is often possible at external examination, the wound morphology
312
complicates the weapon attribution. Whereas pure stab wounds give clues as to the type of blade—and therefore also the type of knife—cuts tend to taper out at both ends and rarely give clues as to the type or the size of blade. Certain serrated cutting edges, as seen, for example, in steak knives will—under ideal conditions—leave telltale “wavy,” or serrated, wound margins. However, the absence of such marks does not exclude such a serrated blade. Waved wound margins are indeed a lucky moment in a forensic pathologist’s daily routine. As the wound morphology belongs to the realm of an external examination, this topic is not discussed further. The direction of cut wounds is usually evident at external examination. However, the question of which type of instrument was utilized and how deep the cut reached may necessitate further investigation. Furthermore, it is not always immediately clear as to what caused death. In addition to the aforementioned telltale wound margins, other structures can help in identifying the involved instrument. If a blade strikes a bone or cartilage, the chances of characteristic marks are rather fair [2,3]. This is true in even badly mutilated, putrefied, or even completely decomposed bodies. In traditional autopsies, greatest care should be taken in dissecting the course of a cut wound of a victim. Careless autopsy technique may give rise to additional scratches and marks. Differentiation between the original marks created by the involved weapon and the also sharp dissecting tool may be complicated or even impossible. Postmortem CT scanning not only serves to display the presence of such deep-lying clues but also aids in the assessment of antemortem and postmortem damage due to difficult or even careless autopsies. Even today, the resolution of the CTs is unfortunately not great enough to display the bone damage in a sufficiently accurate fashion. However, in the light of the rapid improvement of modern CTs, it should only be a matter of time. As in stab wounds, parts of the injury-inflicting instrument may break off and remain in the corpse. Postmortem MSCT can help determine the presence and exact location of such foreign objects. If such an object is extracted, then not only the type of instrument but also which instrument caused the injury can be determined. The depth of a cut may, as mentioned previously, be clear upon mere external examination. However, especially in cases of deeper lesions, the wound bottom may be obscured by blood or by the sheer depth of the cut. In these cases, MSCT can provide additional information as to the depth and the injured structures (Figure D3.4.20, Figure D3.4.21, Figure D3.4.22, Figure D3.4.23, and Figure D3.4.24). MSCT can give clues as to the depth of the cut, but as mentioned in Section D3.4.2 herein, MRI is the method of choice for the visualization of soft-tissue injuries. MRI can display not only the depth of the wound but also accompanying hemorrhages and organ and blood-vessel lesions (Figure D3.4.25, Figure D3.4.26, Figure D3.4.27, Figure D3.4.28, Figure D3.4.29, and Figure D3.4.30). MSCT is, on the other hand, better suited for the detection of intravascular gas and, therefore, a gas, or air, embolism.
© 2009 by Taylor & Francis Group, LLC
The Virtopsy Approach
FIGURE D3.4.20 Autopsy photograph of the right arm of a man who fell through a window during a fight and bled to death. The upper image shows several cuts located in the anterior shoulder region that continue Indian-file-like down toward the inside of the right upper arm (lower image), where a deep cut is visible. The cuts display irregular wound margins due to the thick glass shards of the splintering window.
We conclude that postmortem imaging can be of great value in the assessment of cut wounds: r MSCT delivers information concerning osseous and cartilaginous lesions and therefore further information regarding the depth of a cut, the energy involved, and possibly also the injury-inflicting instrument.
Incident-Specific Cases
313
Chest
Bone Muscle Fat Skin
FIGURE D3.4.21 MSCT 3D reconstruction of the surface of the right shoulder region after virtual removal of the trunk of the case described in Figure D3.4.20. The cuts (arrows) are clearly visible.
FIGURE D3.4.23 MSCT axial image of the right upper arm of the victim shown in Figure D.3.4.10–13. Besides a cut (yellow arrow), gas—in this case air—is seen surrounding the muscles as an indirect radiological sign of an injury (green arrows).
r MSCT can assist in the preautopsy screening of the corpse for foreign bodies, such as parts of knife blades and glass shards, as well as possible bone and cartilage lesions. r MRI is better suited for the investigation of softtissue and blood-vessel injuries.
FIGURE D3.4.22 MSCT 3D reconstruction of the muscles of the right shoulder and upper arm after virtual removal of the torso of the case shown in Figure D3.4.20 and Figure D3.4.21. The muscles display several cuts (yellow arrows), but the veins (green arrows) are intact.
© 2009 by Taylor & Francis Group, LLC
FIGURE D3.4.24 Autopsy photo of the arm of the victim shown in Figure D3.4.20, Figure D3.4.21, Figure D3.4.22, and Figure D3.4.23. The brachial artery (into which a probe has been placed in order to visualize the lesion better) is severed (arrow).
314
The Virtopsy Approach
FIGURE D3.4.25 Autopsy photo of the neck of a victim who committed suicide with neck and wrist cuts. Note the single deep cut (red arrow) accompanied by numerous more superficial cuts (yellow arrows). These superficial lesions, also known as “hesitation cuts,” are indicative of suicide.
D3.4.4 CHOP AND HACKING WOUNDS Chop or hacking wounds are inflicted not only by heavy, not necessarily excessively sharp instruments such as long blades (swords, cutlasses, and machetes) but also by fairly blunt, especially heavy short blades as seen in axes. Compared with injuries inflicted by stabs and cuts, chops are luckily a rarity in the First World. The efficiency in harming and killing has made this cheap and readily available trauma form very popular in genocidal conflicts in the developing world. The devastating conflict between the Hutus and Tutsis in Rwanda (and neighboring countries), with between 500,000 and 1 million casualties, clearly shows that massacre by chopping, like those seen in warfare of bygone days, is by no means over.
FIGURE D3.4.26 T2-weighted axial MRI of the neck of the victim shown in Figure D3.4.25. Note the signal-intense accumulation (yellow arrow), corresponding to blood, in the soft tissues at the right side of the neck.
© 2009 by Taylor & Francis Group, LLC
FIGURE D3.4.27 Autopsy photo of the left wrist of the corpse seen in Figure D3.4.25 and Figure D3.4.26. Two cuts are shown. The depth of these lesions cannot be immediately assessed.
FIGURE D3.4.28 MSCT 3D reconstruction of the wrist seen in Figure D3.4.27. The yellow arrows indicate the cuts; the green arrows point to skin wrinkles of the elderly victim.
Incident-Specific Cases
FIGURE D3.4.29 Depiction of the muscles after virtual removal of the skin of Figure D3.4.28. The two cuts are still visible (arrows), but the wrinkles have disappeared.
FIGURE D3.4.30 3D depiction with MSCT of the bones and tendons of the wrist shown in Figure D3.4.27, Figure D3.4.28, and Figure D3.4.29. No injuries to the tendons in the region of the cuts (circle) are seen.
© 2009 by Taylor & Francis Group, LLC
315
Chops constitute a separate category of sharp trauma, although they share certain traits with the aforementioned cuts. Both cuts and chops consist of a cutting component. However, in cuts, the soft tissue is sliced, thus creating a more or less adaptable and usually abrasion-free wound margin and a relatively circumscriptive lesion. By contrast to cut wounds, chops also have a blunt trauma component. Besides the cutting of soft tissue and to a lesser extent of osseous structures, chops can—depending on the applied force and instrument—crush soft tissue and bones. Severed and amputated extremities are sometimes encountered in this most energy-rich form of (semi) sharp trauma. The wound morphology of chop wounds—namely, a slitlike lesion, often accompanied by an abraded wound margin, destruction of the underlying skeleton, and rarely even bridging in the depth of the wound—can lead to the mistaken diagnosis of a laceration, which is a typical blunt injury. This is especially true in decomposed bodies, where the external examination can be extremely difficult. By virtue of their combined cutting and blunt components, chops are almost always accompanied by cut-like lesions to the adjacent, underlying osseous structures. The identification of gash-like lesions to the bone should immediately give rise to the correct diagnosis (Figure D3.4.31 and Figure D3.4.32). Postmortem MSCT can readily display these telltale cuts or gashes to the underlying bone of a suspected chop lesion. A further advantage is that by this initially no-touch approach, no foreign material (i.e., clues) is displaced into the sometimes huge wounds. Furthermore, MSCT can depict the full extent of the osseous lesions in their correct position. The effort of piecing together a shattered skull after autopsy in order to
FIGURE D3.4.31 MSCT 3D reconstruction of the soft tissues of the head of a homicide victim who was killed by several ax blows roughly three weeks prior to being found. Note the multiple gashes at the right temple. The insert displays the corresponding autopsy photograph.
316
The Virtopsy Approach
FIGURE D3.4.32 MSCT 3D bone reconstruction after virtual removal of the left side of the skull of the victim seen in Figure D3.4.31. Note the large lesion with the telltale gashes to the bone (arrows), indicating several blows. The inflicting weapon was a heavy, rather blunt ax, which led to a combination of sharp (cut) and blunt (blow) trauma.
FIGURE D3.4.33 Skull from a mass grave of the battle of Dornach, Switzerland, from 1499, during which the Swiss broke the siege of the castle Dorneck and eventually defeated the German army. An almost straight cut is seen at the left side of the skull. This type of lesion is typical for a sword blow.
place the pieces in the right anatomical order is considerable and takes a tremendous amount of time and patience. In cases of suspected chops or hacks, postmortem imaging, in this case MSCT, can readily confirm or disprove the initial suspicion by visualizing telltale marks on the underlying bone. Also, foreign objects such as pieces of the blade that broke off during impact may be lodged within the body (Figure D3.4.33, Figure D3.4.34, Figure D3.4.35, Figure D3.4.36, and Figure D3.4.37). Obviously, the identification of such objects, as in cases of stabbing and cutting, is of great importance, as this may give clues as to the type of weapon or even the individual blade used. Postmortem imaging is useful in assessing cases of chops and hacks for the following reasons: r MSCT can identify gashes to the bone easily and therefore can help differentiate chops from lesions due to blunt trauma. This is especially true in cases where the superficial body layers have been damaged, such as in putrefaction and charring. r MSCT can visualize the extent of the damage in a no-touch manner. Thus, as the anatomical structures holding the often shattered bone fragments together are spared, the injury pattern and extent can be appreciated in even severely injured body regions. r MSCT can identify foreign bodies for later extraction and analysis.
© 2009 by Taylor & Francis Group, LLC
FIGURE D3.4.34 Detail of Figure D3.4.33. Note the small object in the cut seen in the insert (arrow).
Incident-Specific Cases
FIGURE D3.4.35 MSCT semitransluscent reconstruction of the skull seen in Figure D3.4.33 and Figure D3.4.34. In this reconstruction, radio-opaque structures such as iron are colored blue. Note the tiny blue object within the cut.
317
FIGURE D3.4.37 Detail of the object seen in Figure D3.4.34, Figure D3.4.35, and Figure D3.4.36. Mass spectroscopy showed that this was indeed corroded iron. Therefore, it is likely that it belonged to a sword and broke off and became lodged in the skull during the blow.
D3.4.5 CONCLUSION MSCT and MRI can contribute—in addition to a thorough external examination—greatly to the medicolegal investigation of deaths due to sharp trauma. They can help assess the wound depth and wound direction, can detect foreign bodies, can detect telltale marks on bones and cartilage, and can help determine the cause of death.
D3.4.6 REFERENCES
FIGURE D3.4.36 MSCT sagittal image of the skull seen in Figure D3.4.33 and Figure D3.4.35. A tiny structure is seen protruding into the cranial cavity from the gash. The region in which this is situated corresponds with the region of the blue object seen in Figure D3.4.35.
© 2009 by Taylor & Francis Group, LLC
Thali M.J., Schwab C.M., Tairi K., et al. 2002. Forensic radiology with cross-section modalities: spiral CT evaluation of a knife wound to the aorta. J Forensic Sci 47:1041–45. Banasr A., Lorin de la Grandmaison G., and Durigon M. 2003. Frequency of bone/cartilage lesions in stab and incised wounds fatalities. Forensic Sci Int 131:131–33. Beaumont E. and O’ Byrne P. 1999. Bone injuries by stab wounds. Report of 19 cases. J Med Legale Droit Med 7–8:581–85. Heuse O. 1982. Beschädigungen der Kleider durch Stichwerkzeuge. Arch Kriminol 170:129–45. Weber W. and Schweitzer H. 1973. Stichversuche an Leichen mit unterschiedlicher kinetischer Energie. Beitr Gerichtl Med 31:180–84. Weber W., Schweitzer H., and Milz U. 1973. Stichdynamik im menschlichen Körpergewebe. Z Rechtsmed 73:295–300. Weber W. and Milz U. 1975. Dynamik manueller Stichversuche. Beitrag zur Aufklärung der “äusseren Stichdynamik.” Z Rechtsmed 75:285–92.
318
The Virtopsy Approach
D3.5 GUNSHOT Stephan A. Bolliger, Beat P. Kneubuehl, and Michael J. Thali
D3.5.1 INTRODUCTION Apart from identification procedures, the field of gunshot injuries is one of the foremost in postmortem forensic radiology. Postmortem radiology serves to locate the projectile and to depict the bullet track and may help in identifying the ammunition and the weapon type utilized. This facilitates the retrieval of the bullet and of potentially important fragments [1]. The retrieval of such foreign bodies is essential, as these objects may display unique rifling characteristics and therefore may help identify the individual weapon used. The knowledge of the bullet course through the body is also of utmost importance, as this may be invaluable when reconstructing the crime scene and the position of the perpetrator and the victim. Postmortem radiology is generally performed using conventional x-ray machines. This technique was incidentally applied almost a century ago in a forensic setting on a gunshot victim [2]. Such conventional x-ray examinations are used in everyday forensic practice in gunshot cases for several reasons: r To detect whether the bullet or part of the bullet has remained in the body r To locate the bullet in the body r To try to determine the type of bullet and thus the ammunition and weapon type in the initial, preautopsy phase of investigations r To try to determine the course of the bullet With the advance of medical technologies such as computed tomography (CT) introduced by Godfrey Hounsfield and Alan Cormack in the early 1970s, new possibilities became available for forensic pathologists. A first CT scan was performed on a victim of a gunshot injury to the head as early as 1977 [3], and authors such as Brogdon [2], Vogel [4], and Donchin [5] have mentioned the usefulness of radiology in forensic medicine. In the past decade, postmortem radiology has grown from a simple autopsy-assisting x-ray to a veritable forensic field of its own. With the invention of spiral computer tomographs, three-dimensional reconstructions of radiological images are possible. Multislice computed tomography (MSCT), which has become an everyday clinical standard, has been implemented in forensic pathology by different groups with promising results, especially regarding the evaluation of gunshot injuries [6–10]. In recent years, these x-ray examinations have been gradually replaced by CT scans [3,7,11–15]. These CT scans have several advantages over conventional x-ray imaging. The latter lacks the three-dimensionality needed to determine the exact location in the body. In order to do this, at least two radiographs have to be performed from different angles. This requires that either the x-ray film plate or the
© 2009 by Taylor & Francis Group, LLC
corpse be moved. This procedure is far more time consuming than performing one rapid MSCT. The advantage of the MSCT is that the forensic pathologist can rotate the 3D image prior to or even during autopsy and thus can gain an exact knowledge of the bullet’s position. Furthermore, MSCT can differentiate different structures with regard to their radioopacity. An x-ray will depict structures of different radiological density in various shades of black, gray, or white. Although this suffices to determine whether foreign objects are within the body, x-rays cannot tell the examiner just how radioopaque the foreign body happens to be. Obviously, not all foreign objects in gunshot incidents are projectiles. Most objects found in gunshot incidents are bone fragments, which are, strictly speaking, not “foreign” but certainly do not belong to soft tissues or organs. Especially if the bullet has passed through an obstacle prior to hitting the victim, different objects may be lodged in the victim. For example, when shooting through a window pane, glass particles of the shattered pane may penetrate the body and be readily confused with bullet fragments in conventional x-rays. With this knowledge prior to autopsy, tedious searches for projectiles in corpses—only to discover that the x-ray image of the assumed bullet fragment was in fact only a piece of bone or some other fragment of higher radioopacity than the surrounding tissue—can be avoided. MSCT can differentiate these foreign objects by virtue of their radioopacity, or Hounsfield unit (HU) characteristics [16].
D3.5.2 GENERAL BALLISTICS Ballistics is defined as the scientific study of projectile motion and encompasses three distinct categories, or flight phases: (1) Internal ballistics describes the projectile’s initial phase of flight within the firearm; (2) external ballistics studies the projectile’s flight in the air; and (3) terminal ballistics analyzes the striking of the target. A subset of terminal ballistics, wound ballistics, describes the motions and effects of a projectile in tissue. Obviously this last field is most important for physicians and forensic pathologists. In order to understand wound ballistics, one must understand the general physical properties of a projectile. A moving projectile possesses kinetic energy. This is determined by the weight and velocity of the bullet as described in the following equation: kinetic energy WV 2 / 2g where W is weight, V is velocity, and g is gravitational acceleration. This formula clearly shows that the velocity plays a greater role than the weight of the bullet. By doubling the velocity, the kinetic energy is increased fourfold, whereas the doubling of the projectile’s weight will only lead to a doubling of the kinetic energy. When the projectile or bullet strikes and penetrates the body, kinetic energy is given to the surrounding tissue. This gives rise to a temporary cavity surrounding the actual bullet course. This temporary cavity, which only lasts for 5–10 seconds, can be considerably
Incident-Specific Cases
larger than the bullet diameter [17,18]. After collapse of the temporary wound cavity, the bullet track, which is generally smaller, remains. The temporary wound cavity is dependent on the tissue structure and on the amount of kinetic energy delivered to the surrounding tissue. Therefore, a projectile that is capable of delivering a large percentage or even all of its kinetic energy to the stricken body will create a large temporary wound cavity and will therefore create more damage than another bullet type. Pistol bullets generally possess less kinetic energy due to their low velocity and therefore can lead to less detrimental temporary wound cavities as, for instance, high-velocity rifle bullets. In the latter case, the temporary cavity may amount to 30 times the bullet diameter. Above a critical velocity [18] of 800–900 m/sec, tissue destruction becomes much more severe. This is the speed encountered in modern assault rifles. The Swiss assault rifle 90 has, for instance, a muzzle velocity of 905 m/sec. However, wound severity depends not only on the kinetic energy a bullet possesses but also on the capacity of delivering this energy to the tissue. If a bullet does not exit the body, it has obviously transmitted all its kinetic energy onto the target. The transmission of kinetic energy to the tissue depends on several other aspects besides the amount of kinetic energy. The yaw of a bullet is the deviation from the long axis from the line of flight. This is comparable to a surfer’s surfboard on a wave crest tumbling toward a beach. The greater the angle of yaw, the greater the loss of kinetic energy to the target is. With increasing flight distance the bullet stabilizes itself, thus decreasing the angle of yaw. This explains why an assault rifle bullet can easily smack right through a head at 100 meters but may disintegrate and remain in the cranium at short distances, such as in suicidal shootings. The shape of a bullet also plays a role in the amount of kinetic energy a bullet loses. A blunt, broad bullet such as a pistol bullet will be slowed down more and therefore will lose more kinetic energy than a long, pointed assault rifle bullet. The deformation properties of a bullet are tremendously important in affecting the degree of kinetic energy loss. A soft tipped or hollow point bullet will deform in the body, thus transmitting a larger percentage of kinetic energy to the tissue. Ideally, such a bullet will mushroom inside the body and therefore will transmit all its kinetic energy, thus bringing about a devastating injury. Regarding properties such as immediate incapacitation due to wound severity or rapid death of the enemy, an ideal bullet should therefore have a high velocity and a maximal angle of yaw that does not affect its accuracy and should be blunt and broad and deform readily inside the body after hitting and penetrating the target.
319
as the bullet striking and entering but not exiting the body. Perforation is reserved for all cases in which the bullet exits. Depending on the distance, gunshot injuries can be divided into three categories: contact, intermediate, and distant. Some authors include a fourth category, the nearcontact wound that fits in neither the definition of contact nor intermediate gunshot wounds. The morphology of gunshot wounds is described exhaustingly in a multitude of forensic textbooks. We therefore refrain from describing these in further detail.
D3.5.4 FORENSIC ASPECTS OF GUNSHOT INJURIES In cases of gunshot injuries, several aspects are of utmost importance. Besides the obvious question of a possible thirdparty involvement, these are as follows: r r r r r r
Weapon handling Firing distance Penetrating or grazing injury Entrance and exit wound Bullet course through the body Gunshot priority in cases of multiple gunshot wounds r Lodging of the bullet or other foreign bodies (i.e., in bullets passing through intermediate targets) in the corpse r Cause of death and vitality of the injuries r Bullet type and size as well as identification of the individual weapon The following sections deal with these aspects in detail. D3.5.4.1 Weapon Handling In contact wounds, the muzzle imprint may give clues as to how the weapon was held against the body upon firing (Figure D3.5.1). This telltale superficial abrasion around the entrance wound can be digitalized with photogrammetrybased 3D surface scanning and matched to the likewise scanned weapon in question (Figure D3.5.2). Under ideal circumstances, not only can a type of firearm be excluded or confirmed, but also the positioning of the weapon onto the skin at the moment of firing can be reconstructed (see also Chapter 2B, Chapter 2D, and Chapter 3D). Thus, the obtained firing position is an important clue in the assessment of contact gunshots. In a suicidal shooting, the victims generally assume a comfortable firing position. A weapon handling that is uncomfortable for the victim must always raise suspicion of third-party involvement. D3.5.4.2 Firing Distance
D3.5.3 GENERAL CLASSIFICATION OF GUNSHOT WOUNDS Gunshot wounds can graze, penetrate, or perforate. Grazing occurs in cases where the bullet strikes the victim extremely tangentially. The resulting abrasion may be easily mistaken for a classic abrasion due to blunt trauma. Penetration is defined
© 2009 by Taylor & Francis Group, LLC
Apart from contact wounds, in which the typical wound morphology proves the distance, the firing distance determination usually belongs to the field of crime scene investigators. Gunpowder tattooing—namely, superficial abrasions due to unburned gunpowder particles striking the skin and,
320
FIGURE D3.5.1 3D photogrammetry-based surface scan of a suicide victim of a contact gunshot wound to the right temple. The entrance wound (green arrow) is as clearly visible as the abrasion inflicted by the recoil spring guide (red arrow) and the housing frame (yellow arrow). The circular objects on the skin are markers for the fusion of surface and internal data.
rarely, the presence of such particles on the skin—may provide a rough firing-distance estimation. Comparison with such a “tattooing” in a firing reconstruction using the original weapon and a reference surface can render the distance estimation more precise. Other methods involve the chemical detection of gunpowder residues on the skin. However,
The Virtopsy Approach
FIGURE D3.5.3 MSCT semitransluscent image of the surface of the right temple of a suicide victim. A circular gray region (red arrows) encompasses the bullet entrance wound (green arrow). This region corresponds to gunpowder residues located under the skin as seen at autopsy (insert), indicating a contact wound.
such an examination requires physical contact with the body, which, in the case of surviving victims, cannot always be performed due to the risk of wound infection. Here again, forensic imaging may be of assistance. Stein et al. [13] showed that gunshot residues can be detected with CT. The metallic particles, especially lead, that are produced by firing are radioopaque and are therefore easily detectable with high-resolution scanners (Figure D3.5.3). In an experimental setting on porcine skin, shots of more than 10 cm could be distinguished from contact gunshots regardless of the type of bullet (i.e., jacketed or solid lead) used. D3.5.4.3 Penetrating or Grazing Injury
FIGURE D3.5.2 Fusion of the data obtained from the 3D photogrammetry-based surface scan of the head and the pistol with the MR data. The superficial lesions seen in Figure D3.5.1 allow for an accurate positioning of the pistol onto the temple at the time of firing, thus permitting a reconstruction of the weapon handling. The bullet course through the brain is clearly seen in the MR image (red line).
© 2009 by Taylor & Francis Group, LLC
Of great importance is the discrimination between penetrating and grazing wounds. The former (Figure D3.5.4 and Figure D3.5.5) may be lethal, whereas the latter are usually harmless. Furthermore, penetrating wounds may result in a bullet lodged in the body that can be examined and can provide clues as to the weapon involved. Additionally, a grazing wound (Figure D3.5.6, Figure D3.5.7, Figure D3.5.8, and Figure D3.5.9) consists basically of a combination of an entry and exit wound, a characteristic that a true penetrating wound does not possess. This difference may lead to confusion regarding the total number of gunshots sustained, as well as the firing position. However, the differentiation of actually penetrating or grazing gunshot wounds is not always easy. The morphology of the wound may be similar in both cases, especially if the penetrating gunshot strikes the victim
Incident-Specific Cases
FIGURE D3.5.4 MSCT surface rendering of a victim who shot himself in the forehead with a pistol. An entrance wound (green arrow) is seen within the typical star-shaped lesion indicating a contact gunshot wound.
tangentially. Putrefaction and maggot infestation can further obscure the wound morphology, making this differentiation more difficult. Postmortem MSCT can, however, readily distinguish a penetrating wound from a grazing wound. By virtue of lead and bone fragments lining the bullet channel, a true penetrating wound can be readily seen in MSCT. Magnetic resonance imaging (MRI), the imaging method of choice, can visualize soft-tissue lesions and therefore can prove a penetrating injury. Obviously, if a lodged bullet is detected by either MSCT or MRI within the body and only one wound is present (Figure D3.5.10, Figure D3.5.11, and Figure D3.5.12), the penetration is clear.
FIGURE D3.5.5 MSCT 3D bone reconstruction of the victim seen in Figure D3.5.4. Besides the entrance defect of the projectile (green arrow) and fracture lines radiating from it, the star-shaped skin defect is seen as a gray shadow (yellow arrows).
© 2009 by Taylor & Francis Group, LLC
321
FIGURE D3.5.6 Autopsy photograph of the vertex of a homicide victim who suffered multiple assault rifle gunshots. The partly torn, partly abraded wound margins with an underlying osseous of this grazing shot can be confused with a laceration due to blunt trauma. In this case, the abraded region (arrows) indicated the area where the bullet struck the head.
D3.5.4.4 Entrance and Exit Wounds The differentiation between entrance and exit wounds is extremely important, as it is absolutely essential for all further case reconstructions. For example, a gunshot to the back with an exit wound at the front of the chest creates a completely different situation from the opposite. Usually, the wound morphology (and in cases of short-distance or contact gunshots also gunpowder residues) suffices to determine whether a wound is due to the entering or exiting of
FIGURE D3.5.7 Autopsy photograph of the vertex of the victim seen in Figure D3.5.6 after removal of the scalp. Upon striking the head, the bullet ploughed its way through the bone.
322
FIGURE D3.5.8 MSCT 3D reconstruction of the skull of the victim seen in Figure D3.5.6 and Figure D3.5.7. The osseous lesions are depicted accurately.
FIGURE D3.5.9 MSCT 3D reconstruction of the skull after virtual removal of the left side of the skull of the homicide case depicted in Figure D3.5.6, Figure D3.5.7, and Figure D3.5.8. Note the inbent bone fragments (arrows), which indicate that the bullet first struck the bone here before ploughing its way onward.
© 2009 by Taylor & Francis Group, LLC
The Virtopsy Approach
FIGURE D3.5.10 Autopsy photograph of the jugulum of a homicide victim who was shot with a revolver. Around the entrance wound (red arrow), multiple tiny, spot-like abrasions are seen (region encircled) that correspond to unburned gunpowder particles striking the skin (so-called gunpowder tattooing). This indicates a close-range firing distance. The circular objects are markers for the fusion of surface and internal data.
FIGURE D3.5.11 MSCT skin reconstruction of the case shown in Figure D3.5.10. The entrance wound (red arrow) is clearly depicted, as are the circular markers.
Incident-Specific Cases
FIGURE D3.5.12 MSCT semitranslucent 3D reconstruction of the thoracic skeleton of the case seen in Figure D3.5.10 and Figure D3.5.11. Objects with a high HU density such as metal are colored blue. Besides the markers (green arrows) and vessel calcifications (red arrows) of this aged victim the bullet is shown (yellow arrow) lying on the right side of the thorax.
a projectile. However, in badly damaged corpses such as charred bodies or putrefaction, neither the wound morphology nor gunshot residue analysis helps in distinguishing an entrance wound. With MSCT, different additional criteria can be used in addressing this concern: Apart from the trail of metal and bone fragments (Figure D3.5.13) along the bullet course, osseous and cartilaginous lesions can provide proof as to the traveling—and therefore entering—direction of the bullet (Figure D3.5.14, Figure D3.5.15, and Figure D3.5.16). In the skull, a cone-shaped beveling helps distinguish an entrance from an exit wound. This phenomenon is described in detail in the following section. D3.5.4.5 Bullet Course through the Body Determining the course of a bullet through the body is extremely important (Figure D3.5.17, Figure D3.5.18, Figure D3.5.19, Figure D3.5.20, Figure D3.5.21, and Figure D3.5.22). First, this can give clues as to how the victim was shot. A cranial-to-occipital sinking bullet course may, for example, be due to a gunshot being fired from above the victim (e.g., from a window or balcony) or against a kneeling or otherwise bent downward victim, to mention just a few possibilities. Obviously, finding a gunshot wound to the back, as opposed to the front, of a victim has immense judicial
© 2009 by Taylor & Francis Group, LLC
323
FIGURE D3.5.13 MSCT sagittal image of the victim seen in Figure D3.5.16. A trail of radioopaque fragments (yellow dashed lines)—in this case bone—leads from the entrance wound to the forehead to the sella turcica. Note the distinct pneumocephalon (X). Due to this pneumocephalon, or rather the sinking of the brain to the posterior cranial cavity due to gravity, the bullet course is not completely in line with the entrance wound.
consequences. A self-defensive action is hardly plausible in a gunshot to a victim’s back. Forensic laypersons often unjustly assume the bullet course to be along the direct pathway between entrance and exit wound. As every experienced forensic pathologist can confirm, this is not always or even rarely the case. Upon striking the body, the bullet can deform and thus, depending on the deformation, can take on a completely “wild” course. A change in tissue texture, especially from a soft tissue to hard bone, may deflect the bullet. The fragmenting of a bullet upon hitting a solid structure may cause two or more pieces to travel at different angles through the air and therefore through the victim. Such wild bullet courses are indeed a challenge for the forensic pathologist and may change the previously assumed cause of death completely. Other, albeit rarer, wild courses are bullet embolisms. In one case, a gunshot to the head traveled through the base of the skull and through the neck and then entered the thoracic aorta (Figure D3.5.23). There, it rested upon cardiopulmonary collapse and gradually sank with the blood to the level of the abdominal aorta. Conventional autopsy procedure in determining the bullet course consists of plane radiographs and a time-consuming layer-by-layer autopsy. Sometimes, a probe is inserted into the wound to determine the general course of the bullet. This blind probing of wounds (i.e., prior to autopsy and trace collection) is not only obsolete but also downright careless. The possible additional gain of information, which would have
324
The Virtopsy Approach
FIGURE D3.5.15 MSCT axial image of the neck of the corpse seen in Figure D3.5.13. Note the thyroid cartilage fragment (yellow arrow) bent inward and bone fragments (green arrows) strewn from the right transverse process to the right side of the nape. Thus, the bullet course (and therefore also a differentiation between entrance and exit wound) can be determined (red arrow).
FIGURE D3.5.14 Autopsy photographs of the front (upper image) and right (lower image) side of the neck of a homicide victim found shot and buried in a forest. The postmortem interval of this moderately decomposed body was around four weeks in cool (about 5n Celsius) soil. Note the oval skin lesions (arrows) on the neck. A differentiation of entrance and exit gunshot wound is not possible as the skin already lacks the epidermis.
been evident at a later stage of the autopsy anyway, does not compensate for the danger of displacing traces such as gunpowder residues and of harming previously uninjured structures. MSCT is capable of depicting the general direction of the bullet by showing which bones were perforated. Often, bone fragments of damaged or shattered bones and bullet fragments “pave” the bullet’s pathway much like a comet’s tail (Figure D3.5.13 and Figure D3.5.24). This bone-fragment trail gives a general impression of the wound direction. However, the state-of-the-art postmortem imaging of bullet pathways in soft tissues is clearly the MRI. As stated before, MRI is capable of visualizing soft-tissue lesions with a high degree of accuracy and is therefore better suited for such examinations. Caution should be exercised, though, in the interpretation of bullet courses through highly mobile
© 2009 by Taylor & Francis Group, LLC
FIGURE D3.5.16 MSCT 3D reconstruction of the thyroid cartilage of the case shown in Figure D3.5.14 and Figure D3.5.15. Note the circular defect (arrow) due to the passing bullet.
Incident-Specific Cases
FIGURE D3.5.17 MSCT 3D bone reconstruction of the spine of a homicide victim who suffered multiple gunshot wounds. Note the small round defect in the sixth thoracic vertebral body (arrow) due to the piercing of a bullet.
FIGURE D3.5.18 Autopsy photograph of the thoracic cavity of the case depicted in Figure D3.5.17. The probe depicts the bullet course through the spine.
© 2009 by Taylor & Francis Group, LLC
325
FIGURE D3.5.19 MSCT coronar reconstruction of a suicidal gunshot to the right temple. Note the small bone fragments (red circle) and the projectile (yellow arrow). The metal components of the bullet give rise to the streak artifacts seen adjacent to the bullet. A slightly lighter bullet track is barely visible (green arrows).
FIGURE D3.5.20 MRI, coronal, T1-weighted image of the brain. The bullet track seen in Figure D3.5.19 is now clearly visible. However, the radioopaque bullet (yellow arrow) and the bone fragments (red circle) are now seen as a signal-silent oval structure.
326
The Virtopsy Approach
FIGURE D3.5.21 Autopsy photograph of the formalin-fixed brain seen in Figure D3.5.19 and Figure D3.5.20. The bullet track (demarcated by green lines) and the bullet (yellow arrow) confirm the aforementioned radiological findings.
organs such as the lungs. The postmortem position of the lungs in an open thoracic trauma, as gunshot wounds to the chest often are, does not correlate to the premortem position, as the lungs may have collapsed after injury.
FIGURE D3.5.23 MSCT axial reformation, soft-tissue window of a homicide victim. Here, a bullet is seen lying right in front of the lumbar spine. Upon autopsy, a bullet was found within the abdominal aorta. The entrance wound of this bullet was located at the head, where it penetrated the skull base and the thoracic aorta and eventually sank down to the level of the abdominal aorta.
D3.5.4.6 Gunshot Priority The term gunshot priority refers to the succession of gunshot injuries to the body. In cases of several gunshot injuries
FIGURE D3.5.22 MSCT coronar reformation, soft-tissue window of the abdomen of a homicide victim. The bullet entered the left side of the abdomen (yellow arrow), transversed the abdomen down toward the right side (dashed lines), and penetrated the right iliac bone, where it remained (red arrow).
© 2009 by Taylor & Francis Group, LLC
FIGURE D3.5.24 MSCT 3D bone reconstruction of a suicidal gunshot to the right temple. The occiput has been virtually removed; the cranium is seen from a dorsal view. The bone fragments (demarcated by a red circle) are clearly visible. The bullet in the left side of the cranial cavity (yellow arrow) has caused streak artifacts due to its metallic composition.
Incident-Specific Cases
sustained from different angles, it is of great reconstructive relevance to determine in which succession these were fired. If a victim displays injuries to the back and the front of the chest, for instance, then the question will arise as to whether the victim was first shot in the back and then, after turning around, into the chest. Differentiating between this scenario and the opposite—namely, a first gunshot to the chest and then into the perhaps fleeing victim’s back—will have obvious judicial consequences. Furthermore, if different persons shot the victim, it may be of relevance to determine which person shot the victim first. This gunshot priority evaluation is not always easy or possible and requires a great amount of experience on behalf of the forensic examiner. Nevertheless, certain scenarios allow for a clear evaluation. In cases of lung perforation, a gunshot priority assessment is possible. In the first phase of multiple gunshot injuries to the chest, the lungs collapse. The subsequent bullet injuries to the lung are therefore more or less in accordance with the thoracic lesions. Therefore, the penetrations that are not in accordance with the postmortem collapsed condition are probably the first injuries, while the opposite is often true for later injuries. Here again, great caution should be exercised in interpreting these possible clues; a hemothorax or tension pneumothorax will obviously deem this general rule incorrect. Another important clue regarding gunshot priority concerns Puppe’s rule, which, described in further detail in the following section, relies upon the fact that a fracture line will not cross a preexisting fracture.
D3.5.4.7 Lodging of the Bullet or Other Foreign Bodies Radiological imaging has proven invaluable in the detection of foreign bodies—the foremost of which are projectiles—in the body. As stated already, the main advantages of MSCT compared with conventional x-ray images lie in their capacity to depict findings three-dimensionally and in different radioopacity gradients (Figure D3.5.25). The three-dimensional depiction of a bullet facilitates the localization of such objects of interest enormously; one flick of a button can locate the projectile in the three-dimensional body accurately. In conventional x-ray examinations, this undertaking required at least two radiographs from different angles. This meant that either the x-ray film plate or the corpse had to be moved. This procedure is far more time-consuming than performing one rapid MSCT. Apart from single bullets within the body, the three-dimensionality of MSCT examinations has also proven to be invaluable in assessing shotgun casualties. It is an almost Herculean undertaking to retrieve or demonstrate the pellets within the body. With x-rays, the general location can be assessed. However, a 3D MSCT reconstruction can depict their distribution within the body accurately, thus making an autoptic assessment superfluous (Figure D3.5.26, Figure D3.5.27, and Figure D3.5.28).
© 2009 by Taylor & Francis Group, LLC
327
FIGURE D3.5.25 MSCT semitransluscent 3D reconstruction of the pelvis of the homicide victim shown in Figure D3.5.23. Metallic structures with a high HU density are colored blue, thus facilitating the detection of the bullet (arrow).
The capability of MSCT to discriminate between different radioopacities is also of great help in pre-autopsy examinations. Conventional x-rays, although undoubtedly sufficient in detecting foreign bodies by virtue of their greater radioopacity than physiologically normal osseous structures, can visualize these only in a black-and-white manner. A further determination of the kind of foreign body (e.g., bullet fragment, glass) is not possible. Unfortunately, it is these other foreign objects that are under certain circumstances invaluable for case reconstructions. If the projectile is fired through an intermediate target such as a window, tiny glass fragments
FIGURE D3.5.26 Classic anteroposterior x-ray of the chest of a man who committed suicide by shooting with a shotgun against his chest. Note the innumerable round shotgun pellets located at the left side of the chest.
328
The Virtopsy Approach
D3.5.4.8 Cause of Death and Vitality of the Injuries Depending on the organs injured in the course of the bullet through the body, a multitude of different causes of death are possible. Of these, death due to craniocerebral trauma, cardiovascular injuries with exsanguination, and gas embolism are the most frequently encountered causes of death. These topics are dealt with in detail in other chapters (see Chapter D2.3, “Vital Reactions and Vital Signs,” Chapter D3.3, “Forensic Nevroimaging,” and Chapter D3.6, “Fatal Hemorrhage in Postmortem Cross-Sectional Radiology”). Another aspect is the so-called vitality of the injuries. It is of obvious judicial relevance whether the sustained gunshot injuries occurred when the victim was alive or already dead. This topic is dealt with in Chapter D2.3. D3.5.4.9 Bullet Type and Size, Identification of the Individual Weapon
FIGURE D3.5.27 MSCT 3D reconstruction of the rib cage of the case shown in Figure D3.5.26 after virtual removal of the posterior portions. The cartilage of the left ribs is destroyed (circle) by the pellets. The shotgun pellets are visible, but streak artifacts disturb the assessment.
may lodge in the body. The identification and retrieval of such fragments with subsequent analysis may give clues to the crime scene, which obviously is not always identical with the location where the corpse was found.
The crime scene investigators often need information as to the type of bullet involved as soon as possible in order to include or exclude different weapon types. Usually this meant waiting until the projectile or fragments of it were retrieved at autopsy. With MSCT, the size and form of the more or less intact projectile can be assessed accurately prior to retrieval. Furthermore, under ideal conditions the type of the bullet with respect to it being jacketed or not, and even the metal used for the jacket can be gained with MSCT [16]. Therefore, the crime scene investigators have additional information regarding the weapon involved even before autopsy. Due to the limited resolution of MSCT today, the rifling of the bullet cannot be assessed accurately enough in the body, making the retrieval necessary. This retrieval of the whole bullet or bullet fragments is of utmost importance, as the rifling on the projectile may lead to the identification of the individual weapon and therefore lead to the perpetrator. This undertaking is enormously facilitated by postmortem radiology, especially MSCT, which suffices in locating the projectile within the body accurately and therefore allows for a minimally invasive extraction before or even instead of an autopsy.
D3.5.5 HEAD INJURIES
FIGURE D3.5.28 MSCT semitransluscent 3D reconstruction of the rib cage seen in Figure D3.5.26 and Figure D3.5.27. HU-dense objects such as the lead shotgun pellets are colored blue, facilitating their detection.
© 2009 by Taylor & Francis Group, LLC
Although the head compromises only a small percentage of the human body, head injuries due to gunshot wounds are a frequent finding in forensic practice and therefore warrant a separate section. Due to the fact that head injuries often lead to a rapid death, this small body region is frequently targeted in suicidal as well as homicidal gunshots. Gunshots to the head leave a multitude of traces and clues that are also generally easily detectable with postmortem imaging procedures described as follows. The wound morphology gives clues as to the distance and—in cases of contact wounds—also of the weapon type. By virtue of the rigid skull, the presence of other clues is a rule. If a bullet strikes a skull tangentially, so-called gutter
Incident-Specific Cases
wounds arise [19]. These may be confined to the outer table or even the inner table. In such cases, low-velocity bullets can slip between the skull and the tough tissue of the scalp to a completely distant part of the head. Without prior radiography, these gunshots can cause certain moments of surprise at autopsy. Indeed, the localization of such ultimately flattened bullets between the scalp and the skull is difficult, as they frequently drop out of their position when the scalp is drawn off the skull at autopsy. Postmortem MSCT is able to locate such projectiles rapidly and therefore helps to avoid time-consuming searches for possibly paper-thin bullets. If the bullet strikes the head at a shallow angle, it may split into two fragments while penetrating the bone. One part can then glance off the skull and leave only a superficial bone lesion, while the other penetrates the skull. This gives rise to the keyhole-shaped wound [20]. A more classical form of gunshot injury to the head is the penetrating and often perforating trauma. Here, the bullet typically enters the skull, traverses through the brain, and may or may not—depending on the energy of the bullet—exit the skull. Obviously, the differentiation of exit and entrance wound is immensely important for reconstructive purposes. The skull, by virtue of it consisting to a large extent of more or less flat bones, makes the differentiation between entrance and exit wounds fairly easy. When a bullet passes through these flat bones in a more or less perpendicular angle, the result is a round or round-oval defect in the bone. This is true for exit and entrance wounds. On the opposite side of the bone, the inner table is beveled away from the bullet impact site, thus creating a cone-like appearance. The tip of the cone points toward the gun and therefore enables a differentiation between entrance and exit wound. In entrance wounds, the inner table is beveled out, called inward beveling (Figure D3.5.29). Exit wounds display the opposite: Here, it is the outer table that is beveled out, called outward beveling (Figure D3.5.30). The skin-wound morphology with regard to entrance and exit wound may be difficult for the naked eye to interpret due to a number of reasons (e.g., extensive lesions, putrefaction, insect and other animal involvement). However, postmortem imaging is capable of demonstrating the beveling features of the bone with sufficient accuracy as to determine exit and entrance wounds. Another feature that gunshot wounds to the skull often possess is the presence of secondary fractures. As the skull is a rigid structure, a sudden increase of the intracranial pressure as in penetrating gunshot injuries to the head may give rise to a secondary fracturing of the skull at its most fragile areas (Figure D3.5.31 and Figure D3.5.32). Such secondary fractures depend on the firing distance and the bullet’s kinetic energy. Contact wounds give rise to extensive secondary fractures due to the gas produced by the discharge entering the head (Figure D3.5.33). The most extreme form of secondary fracturing is the socalled Kroenlein gunshot. Here, a gunshot with high-velocity ammunition is applied directly to the head. The result of this explosion-like force to the head is a complete destruction of the skull with cerebral extenteration. The forensic relevance of such secondary fractures is the so-called gunshot priority
© 2009 by Taylor & Francis Group, LLC
329
FIGURE D3.5.29 MSCT coronar reformation of a suicidal gunshot to the vertex. Note the inward beveling of the entrance defect indicated by red lines, the pneumocephalon and the trail of bone, and bullet fragments to the skull base (demarcated by yellow lines).
FIGURE D3.5.30 MSCT axial image of the head of a suicidal gunshot to the forehead. Besides the pneumocephalon (X), a trail of bone fragments (demarcated by yellow lines) leads to the occiput where an exit wound with an external beveling (indicated by red lines) is seen.
330
The Virtopsy Approach
FIGURE D3.5.31 MSCT 3D reconstruction of the skull of a 5-year-old homicide victim. One bullet entered the forehead and passed slightly upward to the left side of the occiput. There, it exited the skull. Note the tiny bone fragments at the exit wound and the fractures due to hydrodynamic forces.
FIGURE D3.5.33 MSCT 3D skull reconstruction of a man who shot himself while playing Russian roulette. Note the extensive damage of skull around the entrance defect (green line) and the exit wound seen on the other side of the skull (red arrow). The skull literally burst open due to the hydrodynamic effect of the 9 mm bullet.
FIGURE D3.5.32 MSCT 3D reconstruction of the skull of the case seen in Figure D3.5.31. Here, the second bullet course is depicted: It struck the lower jaw tangentially from the front (dashed arrow) and exited the head just behind the right ear.
in multiple wounds to the head. Fracture lines will not pass preexisting fractures, as the strain on the bone—the essential prerequisite for fracture formation—ceases at the preexisting fracture. This rule, also known as Puppe’s rule after its first describer, allows for the determination of the gunshot priority in multiple wounds to the head. Obviously, such telltale fractures are not detected at external examination. These secondary fracture lines are clearly visible in MSCT. Therefore, the gunshot priority can be assessed before autopsy. As the brain is almost completely encapsulated by the skull, a bullet will invariably have to penetrate the skull in order to create cerebral injuries. The chances for the bullet to leave a trail of lead, steel, or bone fragments are therefore fairly great. Due to this fragment trail, MSCT can depict the bullet course through the brain easily. However, caution should be exerted in overestimating the accuracy of such a bullet course through the brain; it does not necessarily have to correspond exactly to the bullet course through the head. If the injury is survived for a certain amount of time, the injured brain may swell. This cerebral edema can be onesided and therefore lead to a shift of the cerebral structures and therefore of the seen bullet course. Oppositely, a shrinking of the brain, as encountered in a pneumocephalon will alter the course of the bullet through the brain compared with the direct path between entrance and exit defect. Postmortem MRI is not effective in displaying hard foreign particles, as these generally do not contain water. Therefore, bullets, glass, and (to some extent) bone will remain as signal-
© 2009 by Taylor & Francis Group, LLC
Incident-Specific Cases
silent structures. The cerebral tissues and their respective damage (e.g., contusion, hemorrhages) are well distinguishable. Occasionally, even the bullet course can be depicted.
D3.5.6 CONCLUSION We believe that postmortem imaging—ideally a combination of MSCT and MRI—is a helpful adjuvant to the state-of-theart classic forensic examination of ballistic trauma. MSCT can differentiate between entrance and exit wounds, can detect bullets and other foreign bodies in the corpse, can provide information on the injured organs and therefore the cause of death, and can give a general overview of the bullet course. Furthermore, MSCT can address the question of gunshot priority in multiple wounds. MRI gives, besides a general impression of the bullet course, detailed information on soft-tissue injuries. The cause of death is therefore more accurately detailed in MRI than in MSCT examinations. However, MRI does not depict hard objects such as the bullet or other foreign particles sufficiently.
331
12. Schumacher M., Oemichen M., Konig H.G., et al. 1985. Computer tomographic studies on wound ballistics of cranial gunshot injuries. Beitr Gerichtl Med 43:95–101. 13. Stein K.M., Bahner M.L., and Merkel J. 2000. Detection of gunshot residues in routine CTs. Int J Legal Med 114:15–18. 15. Thali M.J., Watzke O., and Kachelriess M. 2001. Bullets and metal artefacts—state of the art. Rechtsmedizin 11:193. 16. Jackowski C., Lussi A., Classens M., et al. 2006. Extended CT scale overcomes restoration caused streak artifacts—3D color encoded automatic discrimination of dental restorations for identification. JCAT 30:510–13. 17. French R.W. and Callender, G.R. 1962. Ballistic characteristics of wounding agents. In Wound Ballistics, ed. Beyer, J.C. Washington, DC: Superintendent of Documents, U.S. Government Printing Office. 18. Callender G.R. and French R.W. 1935. Wound ballistics: studies in the mechanism of wound production by rifle bullets. Mil Surg 77:177–201. 19. La Garde L.A. 1916. Gunshot Injuries. New York: William Wood and Co. 20. Dixon D.S. 1982. Keyhole lesions in gunshot wounds of the skull and direction of fire. J Forensic Sci 27:555–66.
D3.5.7 REFERENCES 1. Di Maio V.J.M. 1999. Gunshot Wounds: Practical Aspects of Firearms, Ballistics, and Forensic Techniques. Boca Raton, FL: CRC Press. 2. Brogdon B.G. 1989. Forensic Radiology. Boca Raton, FL: CRC Press. 3. Wullenweber R, Schneider V., and Grumme T. 1977. [A computer-tomographical examination of cranial bullet wounds]. Z Rechtsmed 80:227–46. 4. Vogel H. 1997. Gewalt im Röntgenbild: Befunde bei Krieg, Folter und Verbrechen. Echomed 41:13–42. 5. Donchin Y., Rivkind A.I., Bar-Ziv J., et al. 1994. Utility of post-mortem computed tomography in trauma victims. J Trauma 37:552–55. 6. Thali M.J., Yen K., Schweitzer W., Vock P., et al. 2003. Virtopsy, a new imaging horizon in forensic pathology: virtual autopsy by postmortem multisclice computed tomography (MSCT) and magnetic resonance imaging (MRI)—a feasibility study. J Forensic Sci 48:386–403. 7. Thali M.J., Yen K., Vock P., et al. 2003. Image-guided virtual autopsy findings of gunshot victims performed with multi-slice computed tomography (MSCT) and magnetic resonance imaging (MRI), and subsequent correlation between radiology and autopsy findings. Forensic Sci Int 138:8–16. 8. Thali M.J., Schweitzer W., Yen K., et al. 2003. New horizons in forensic radiology: the 60-second digital autopsy— full body examination of a gunshot victim by computed tomography. Am J Forensic Med Path 24:22–27. 9. Levy A.D., Abbott R.M., Mallak C.T., et al. 2006. Virtual autopsy: preliminary experience in high-velocity gunshot wound victims. Radiology 240:522–28. 10. Harcke H.T., Levy A.D., Abbott R.M., Mallak. 2007. Autopsy radiography. Digital radiographs (DR) vs multidetector computed tomography (MDCT) in high-velocity gunshot-wound victims. Am J Forensic Med Path 28:13–19. 11. Thali M.J., Kneubuehl B.P., Bolliger S.A., et al. 2007. Forensic veterinary radiology: ballistic-radiological 3D computertomographic reconstruction of an illegal lynx shooting in Switzerland. Forensic Sci Int 171:63–66.
© 2009 by Taylor & Francis Group, LLC
D3.6 FATAL HEMORRHAGE IN POSTMORTEM CROSS-SECTIONAL RADIOLOGY Emin Aghayev, Michael J. Thali, and Peter Vock In forensic medicine, fatal hemorrhage belongs to a relatively frequent cause of death. It can be caused by either blunt or sharp force as well as by a number of diseases, such as peptic ulcer and varicosis. A distinct hemorrhage leads primary to a hypovolemia and hypotension, which then cause hypoperfusion and hypoxia resulting in heart failure and brain hypoxia. Nevertheless, fatal hemorrhage is held as a separate cause of death in forensic medicine. Determination of loss of blood at autopsy is often difficult. In classic traditional autopsy, no objective quantitative system of accurately measuring the blood volume in the corpse has been developed, and thus, as in many other areas of forensic medicine, the pathologist has to use simple subjective means of making this determination, using his or her personal experience and power of observation. The determination of fatal hemorrhage in forensic medicine is based upon three major premises. First is the extent of livor mortis, which the forensic pathologist assesses already by external inspection. A decrease in the extent of livor mortis is deemed as a foremost premise for fatal hemorrhage. Hereunto comes the finding of extravasal blood in the body. The estimation of the volume of extravasal blood (e.g., in the thoracic or abdominal cavities) at autopsy may serve as a further premise for fatal hemorrhage. Additionally, subjective assessment of the color of internal organs regarding their paleness at autopsy may provide a supplementary clue for blood loss. Despite the selection of clues for fatal hemorrhage, forensic pathologists know that all these arguments are weak. The low sensitivity of the extent of livor mortis was mentioned by Knight [1] in his textbook as follows: “The
332
The Virtopsy Approach
phenomenon (livor) appears at a variable time after death, indeed it may not appear at all, especially in infants, old people, those with anemia. It may be so faint as almost to escape detection.” Estimation of the volume of blood loss becomes difficult when the blood is pooled in hematomas or ecchymoses and almost impossible when the blood has disappeared at the crime scene. The paleness of vascular organs is a relative criterion, and it may be caused by etiologies other than acute bleeding such as chronic anemia; furthermore, paleness may exist in nondepending areas without bleeding. Moreover, in contrast to any clinical or radiologic department there is no verification of the conclusion of forensic autopsy because it is the last step of investigation. The recent implementation of radiological cross-sectional methods in forensic medicine has brought a number of benefits for forensic investigation, among which is the possibility of noninvasive objective documentation of the findings [2–4]. It was supposed that reduction of the diameter of the major body vessels in postmortem imaging might be a sign for fatal hemorrhage [4]. The idea of measuring the cross-sectional area of major blood vessels in postmortem multidetector computed tomography (MDCT) and magnetic resonance imaging (MRI) was born after the recent development of postmortem imaging. Experienced eyes of radiologists observed an outstanding difference in the cross-sectional areas of major blood vessels between living patients and in corpses (Figure D3.6.1, Figure D3.6.2, Figure D3.6.3, Figure D3.6.4, and Figure D3.6.5, Figure D3.6.6, Figure D3.6.7, and Figure D3.6.8).
D3.6.1 CLINICAL RADIOLOGICAL EXPERIENCE The association of hypovolemia and hypotension with flattened vessels, especially the cava inferior vein in abdominal
AA
CIV
ABA
FIGURE D3.6.2 Axial MDCT images of a living person on the level just below the liver. Note the plump-filled cava inferior vein (CIV) and abdominal aorta (ABA).
CT scans, has been used in clinics since the late 1980s [5–8]. A strong correlation of hypovolemia with flattening of the inferior vena cava at multiple levels on CT images in a patient who had suffered a substantial blunt abdominal trauma was documented by Jeffrey and Federle [5]. In their retrospective study of 100 trauma patients, the cava flattening was observed in seven cases. Six of the seven had major hemorrhage. A control group of 100 abdominal CT scans showed no flattening in 98 cases. Taylor et al. postulated that a flat cava may indicate an impending cardiovascular collapse [6]. More lately, Sivit et al. described the hypoperfusion complex in 27 of 1,018 children [7]. Cava flattening was observed in all 27 cases. Mirvis et al. reported the presence of a flattened inferior vena cava on abdominal CT in 10 of 13 shock patients after blunt abdominal
MPA
CSV AA
MPA
CSV
RPA DA RPA
FIGURE D3.6.1 Axial MDCT images of a living person on the level just above the coronary orifices. Note the plump-filled cava superior vein (CSV), main pulmonary artery (MPA), right pulmonary artery (RPA), ascending aorta (AA), and descending aorta (DA).
© 2009 by Taylor & Francis Group, LLC
DA
FIGURE D3.6.3 Axial MCDT image of a deceased just above the coronary orifices. Note the less plump-filled marked major vessels: CSV: cava superior vein, MPA: main pulmonary artery, RPA: right pulmonary artery, AA: ascending aorta, and DA: descending aorta.
Incident-Specific Cases
333
CSV
AA MPA DA P A R
ABA CIV
FIGURE D3.6.4 Axial MCDT image of a deceased just below the liver. Note the less plump-filled marked CIV and ABA.
trauma [8]. In contrast, Eisenstat et al. reported abdominal CT scans of 500 nontraumatic patients [9], where the flat cava was observed in 14% of all patients and only the minority (30%) of them had a hypotension or evidence of hypovolemia. On postmortem CT examinations, Thali et al. first described a reduction of the diameter of the abdominal aorta, the so-called aortic collapse sign, which the authors supposed may be a reproducible radiological sign of massive or even fatal hemorrhage [4].
FIGURE D3.6.6 Axial MDCT image on the level of just above the coronary orifices of a case who died after a vehicle accident. The CSV, MPA, and RPA are completely collapsed and AA and DA are partially collapsed.
AA
MPA
CSV DA P A R
D3.6.2 POSTMORTEM IMAGING FINDINGS IN CASES WHO DIED DUE TO FATAL HEMORRHAGE Postmortem imaging findings within the cardiovascular system were recently studied in 19 cases who died due to fatal
FIGURE D3.6.7 Axial MDCT image on the level of just above the coronary orifices of another case who died after a fall from height. Note the completely collapsed CSV and AA with gas as well as collapsed MPA, RPA, and DA.
1: distance 5.8 mm
AA MPA
CSV
CIV ABA
1 DA
FIGURE D3.6.5 Axial MDCT image on the level of just above the coronary orifices of a case who died after a fall from height. Note the collapsed CSV with a central catheter, MPA, AA, and DA. Measurement of the diameter of the RPA (distance 1) is 5.8 mm (see also Figure D3.6.9).
© 2009 by Taylor & Francis Group, LLC
FIGURE D3.6.8 Axial MDCT image on the level just below the liver of a case who died due to an open craniocerebral trauma. Both CIV and ABA are collapsed.
334
The Virtopsy Approach
$# ' &
&
"# '
&
$ %'
!
of the four heart chambers if specificity higher than 79% is wished. For the remaining measurements MDCT appears as a proper method. The case selection for the described study was based on the autopsy protocols that means on the protocols with all those limitations mentioned in the introduction. There is a current need for postmortem imaging studies on fatal hemorrhage to document benefits and limitations of both autopsy and cross-sectional imaging methods. Quantitative postmortem cross-sectional imaging is a promising objective noninvasive method for assessing the question of fatal hemorrhage in forensic medicine currently as a supplement to today’s autopsy.
D3.6.3 REFERENCES FIGURE D3.6.9 The diagram shows three-step algorithm for the differentiation of fatal hemorrhage from other causes of death. Steps 1 (RPA) and 2 (MPA) are MDCT based, whereas step 3 (right atrium, or RAT) requires MRI data-based volume measurement
hemorrhages and 46 cases with different causes of death. According to the published work, in cases who died due to fatal hemorrhage half of the major body vessels—such as the aorta, main pulmonary artery, right pulmonary artery, portal vein, cava superior, and inferior veins—were observed to be collapsed (Figure D3.6.5, Figure D3.6.6, Figure D3.6.7, and Figure D3.6.8) [10]. Furthermore, the finding of a collapsed superior vena cava, main pulmonary artery, or right pulmonary artery—together or alone—is a clear clue to fatal hemorrhage, and a collapsed inferior vena cava or ascending or descending aorta also point to a fatal hemorrhage as a combined or single cause of death (Figure D3.6.5, Figure D3.6.6, and Figure D3.6.7) [10]. In addition to the aforementioned findings, an algorithm for the radiological differentiation between cases who died due to fatal hemorrhage and cases who died due to other causes of death with 100% sensitivity was proposed (Figure D3.6.9). The specificity of 95% was reached whenever the greatest diameter of the right pulmonary artery measured on the axial CT images was ≤ 6 mm; the crosssectional area of the main pulmonary artery also measured on axial CT images was ≤ 130 mm2; and the volume of the right atrium estimated using MRI data was equal to or less than 13 cm3 (Figure D3.6.9). When only the CT- or MRI-derived parameters are used, the reached specificity is 79%. The availability of only one of these devices for examination is much more probable because of their high cost. In contrast to MRI, CT is a less expensive cross-sectional radiological method with faster scanning times, which is also more available in forensic medicine. Today, approximately 15 forensic institutes worldwide use a CT scanner for examination of forensic bodies. Regrettably, the limited soft-tissue density differentiation of MDCT [3] does not allow an accurate measurement of the volume of the heart chambers as well. For this reason, the use of MRI is recommended for examination of the volume
© 2009 by Taylor & Francis Group, LLC
1. Knight B. 1991. Forensic Pathology. London: Arnold. 2. Aghayev, E., M. Thali, C. Jackowski, et al. 2004. Virtopsy— fatal motor vehicle accident with head injury. J Forensic Sci 49:809–13. 3. Thali, M. and P. Vock. 2003. Role of and techniques in forensic imaging. In: Forensic Medicine: Clinical and Pathological Aspects, ed. J. Payen-James, A. Busuttil, and W. Smock, 731–45. London: Greenwich Medical Media. 4. Thali, M.J., K. Yen, W. Schweitzer, et al. 2003. Virtopsy, a new imaging horizon in forensic pathology: virtual autopsy by postmortem multislice computed tomography (MDCT) and magnetic resonance imaging (MRI)—a feasibility study. J Forensic Sci 48:386–403. 5. Jeffrey, R.B., Jr., and M.P. Federle. 1988. The collapsed inferior vena cava: CT evidence of hypovolemia. AJR 150:431–32. 6. Taylor, G.A., M.E. Fallat, and M.R. Eichelberger. 1987. Hypovolemic shock in children: abdominal CT manifestations. Radiology 164:479–81. 7. Sivit, C.J., G.A. Taylor, D.I. Bulas, D.C. Kushner, B.M. Potter, and M.R. Eichelberger. 1992. Posttraumatic shock in children: CT findings associated with hemodynamic instability. Radiology 182:723–26. 8. Mirvis, S.E., K. Shanmuganathan, and R. Erb. 1994. Diffuse small-bowel ischemia in hypotensive adults after blunt trauma (shock bowel): CT findings and clinical significance. AJR 163:1375–79. 9. Eisenstat, R.S., A.C. Whitford, M.J. Lane, and D.S. Katz. 2002. The “flat cava” sign revisited: what is its significance in patients without trauma? AJR 178:21–25. 10. Aghayev, E., M. Sonnenschein, C. Jackowski, et al. 2006. Postmortem radiology of fatal hemorrhage: measurements of cross-sectional areas of major blood vessels and volumes of aorta and spleen on MDCT and volumes of heart chambers on MRI. AJR 187:209–15.
D3.7 STRANGULATION Lars Oesterhelweg and Michael J. Thali Strangulation is an important and common entity of the causes of death in medicolegal investigation. There are three types of strangulation: hanging, ligature strangulation, and manual strangulation (Table D3.7.1).
Incident-Specific Cases
335
TABLE D3.7.1 Types of Strangulation Hanging Ligature strangulation Manual strangulation
Pressure to the neck by the own body weight in an instrument Pressure to the neck by an instrument and another force Pressure to the neck by the hands of another person
While today death from hanging in the majority of cases is suicidal, ligature strangulation may be either homicidal or suicidal, and manual strangulation is in nearly all cases homicidal. In general, due to the dynamic events that take place between the victim and the perpetrator, the injuries are more severe than those in suicide, which results in findings that are more noticeable on forensic examination. Characteristic in all three forms is an external pressure on the neck with subsequent closure of the blood vessels in the neck, and death is caused by cerebral hypoxia (Table D3.7.2). The major blood flow is in the carotid arteries and the jugular veins. While the pressure necessary to occlude the jugular veins is relatively low, the pressure on the carotid arteries is approximately 5 kg (11 lb) for occlusion [1]. This relatively low weight pressure explains that suicidal hanging is possible in a seated or kneeling position. Closure of the carotid arteries causes unconsciousness in an average of 10 sec due to hypoxia of the brain. Irreversible cerebral damage may occur in a period of more than 4 min of occlusion. Partial closure of the arteries and the jugular veins will lead to a rapid rise of the venous blood pressure in the head. This is represented by the classic external signs of strangulation—congestion, cyanosis, edema, and petechial hemorrhages—above the line of constriction [2]. In the event that the circular pressure to the neck is more than 5 kg (11 lb), these classic signs may be missing due to the lack of congestion. In most manual and ligature strangulation cases, the classic external signs are present; in hanging, the signs may be absent. Two types of hanging— white and blue—can be described. In the white death from hanging, the carotid arteries are occluded immediately, and the blood flow persists as well in the carotid arteries and in jugular veins. In these cases, the loss of consciousness will be quick by hypoxia, and the face remains pale. For the occlusion of the arteries, normally a free hanging of the
TABLE D3.7.2 Major Blood Vessels in the Neck to and from the Brain
body with a symmetric position of the rope and the knot at the highest point in the neck is described. In other cases where the rope is in an asymmetric position or the feet have contact with the ground, the classic external signs are present due to a partial occlusion of the arteries. These signs of congestion lead to a blue discoloration of the skin above the line of constriction [3]. In addition to the findings caused by congestion, on external examination bruising from the assailant and fingernail abrasions from the victim can often be found in manual strangulation, while in ligature strangulation and hanging a strangulation mark caused by imprint of the rope or strangulation instrument is visible [2,3]. If an autopsy is performed, it may reveal typical strangulation findings such as soft-tissue hemorrhages or fractures of the laryngeal skeleton. In clinical radiology, experience with computed tomography (CT) and magnetic resonance imaging (MRI) of these injuries is minimal. CT examinations of the neck and throat in cases of strangulation have been described by several authors, while there are only rare reports of MRI findings [4–14]. The main radiological findings in the neck and cervical spine region were thyroid and hyoid fractures in addition to ruptures of the ligaments of the cervical spine and fractures of the transverse processes. Even in plain radiographs and CT images in a case of attempted suicide by hanging, an extensive hemorrhage of the soft tissues could be observed [8].
D3.7.1 STRANGULATION BY HANGING The two primary findings in hangings—a strangulation mark and subcutaneous desiccation (i.e., soft-tissue thinning as a result of tissue fluids being driven out by mechanical compression)—are radiologically detectable by both MRI and CT in hanging cases (Figure D3.7.1). Three-dimensional postprocessing of the multislice CT (MSCT) data allows the demonstration of the strangulation mark characteristics in each case (e.g., exact position and width, and patterning). Excellent concordance between the autopsy and radiology findings is found in cases presenting with subcutaneous hemorrhage, lymph node hemorrhage, platysmal hemorrhage, and fracture of the laryngeal skeleton and in cases showing congestion and hemorrhage of the salivary glands [12,14]. In addition, intramuscular hemorrhage can be depicted by imaging (MRI). While small hemorrhages in the laryngeal muscles can be visualizable only infrequently by MRI and MSCT, fractures of the osseous or cartilageous structures of the larynx can be found in most cases [14,15] (Figure D3.7.2).
Major Blood Vessels to the Brain
Major Blood Vessels from the Brain
D3.7.2 MANUAL STRANGULATION
Carotid arteries Vertebral arteries Small spinal arteries
Jugular veins Vertebral venous plexus Deep cervical veins
In deceased strangulation victims, the primary finding of subcutaneous hemorrhage as well as intramuscular hemorrhage is detectable detected by both CT and MRI (Figure D3.7.3). All other autopsy findings, such as subcutaneous desiccation,
© 2009 by Taylor & Francis Group, LLC
336
The Virtopsy Approach
A
B
FIGURE D3.7.1 Primary signs in strangulation by suicidal hanging: (A) Coronal short tau inversion recovery (STIR) MR image showing subcutaneous desiccation in the strangulation mark area (small arrows) and a deep impression of the skin in the lateral neck parts (larger arrows). Subcutaneous desiccation underneath the strangulation mark results from compression of the skin and subcutaneous layers due to the strangulation mechanism and is seen as a hypointense signal in this STIR MR image. (B) At autopsy, band-like subcutaneous desiccation is present in the region of the strangulation mark.
lymph node hemorrhage (Figure D3.7.4), congestion and hemorrhage of the submandibular glands (Figure D3.7.5), and fractures of the laryngeal skeleton are very visualizable at imaging. The macroscopic and microscopic hemorrhage of the lymph nodes is present in the majority of the manual A
A
B B
D
C
D
FIGURE D3.7.2 Osseous injury in a case of suicidal hanging: (A) MSCT showing a fracture of the left hyoid bone (small arrow). The strangulation mark appeared as a deep impression of the skin on both sides of the neck (large arrows). (B) 3D reconstruction of MSCT data clearly depicting the deviation at the fracture site. (C) Left hyoid bone specimen after the maceration procedure. Even after maceration of the bone, the fracture was hardly detected (arrows). (D) Electron microscopic examination (r 29, 4 kV) demonstrating the fracture line through the entire diameter of the hyoid bone.
© 2009 by Taylor & Francis Group, LLC
C
FIGURE D3.7.3 Primary signs in manual strangulation: subcutaneous and intramuscular hemorrhage: (A) Contusion in the subcutaneous fatty tissue above the right mandibular angle. This type of injury is characteristic of manual strangulation and results from pressure on the mandibular angles during the assault. (B) The same region at autopsy, after incision of the subcutaneous fatty tissue. The lobules of fat tissue were contusioned and hemorrhagic. (C) Intramuscular hemorrhage (arrows) depicted in the left sternocleidomastoid muscle by axial T2-weighted fat-saturated MRI. The muscle shows hyperintensities and swelling in comparison with the right side. (D) Moderate intramuscular hemorrhage at autopsy (arrows).
Incident-Specific Cases
337
indicate that lymph node hemorrhage is also a specific diagnostic sign of strangulation indicating the pathomechanism of congestion. The limited resolution of the imaging methods might be the reason for the failure to detect small hemorrhage, which were detectable at autopsy. Ongoing clinical research (e.g., using high-Tesla MR scanners, microscopic imaging) will improve the spatial resolution and therefore will facilitate the diagnosis of tiny alterations in the future [14].
A
B
C
D3.7.3 REFERENCES
FIGURE D3.7.4 Lymph node injury following manual strangulation: (A) A hyperintense lymph node was detected on the left side of the neck (large arrow) on a coronal T2-weighted fat-saturated MR image. The left shoulder showed contusioned, hyperintense subcutaneous fatty tissue (small arrows). (B) Forensic-pathological correlation at autopsy. The lymph nodes appear dark red (arrow). (C) Histological specimen (r 25) demonstrating extensive hemorrhage (large arrows) in the lymph node. A layer of erythrocytes in the connective tissue surrounding the lymph node (small arrows) confirmed the traumatic genesis of the finding.
strangulation cases, including the living patient (see Chapter D3.11, “Clinical Forensic Imaging”), and is often eye-catching in the pre-autopsy radiological examination. This sign— almost unknown in classic forensic medicine [16]—may
1. DiMaio, V. J. and D. DiMaio. 2001. Forensic Pathology, 2d ed. Boca Raton, FL: CRC Press. 2. Saukko, P. and B. Knight. 2004. Knight’s Forensic Pathology. London: Arnold. 3. Brinkmann, B. 2004. Ersticken. In Handbuch gerichtliche Medizin, Band 1, ed. B. Brinkmann and B. Madea, 761–94. Berlin: Springer-Verlag. 4. Ikenaga, T., M. Kajikawa, H. Kajikawa, et al. 1996. Unilateral dissection of the cervical portion of the internal carotid artery and ipsilateral multiple cerebral infarctions caused by suicidal hanging: a case report. No Shinkei Geka 24:853–58. 5. Kiani, S.H. and D. C. Simes. 2000. Delayed bilateral internal carotid artery thrombosis following accidental strangulation. Br J Anaesth 84:521–24. 6. Linnau, K. F. and W. A. Cohen. 2002. Radiologic evaluation of attempted suicide by hanging: cricotracheal separation and common carotid artery dissection. AJR 178:214. 7. Lupetin, A. R., M. Hollander, and V. M. Rao. 1998. CT evaluation of laryngotracheal trauma. Semin Musculoskel Radiol 2:105–16. 8. Meglin, A. J., J. F. Biedlingmaier, and S. E. Mirvis. Threedimensional computerized tomography in the evaluation of laryngeal injury. Laryngoscope 101:202–07. 9. Noguchi, K., Y. Matsuoka, K. Hohda, J. Katsuyama, and S. Nishimura. 1992. A case of common carotid artery stenosis due to hanging. No Shinkei Geka 20:1185–88.
A
C
B
FIGURE D3.7.5 Submandibular gland alterations in a case of manual strangulation: (A) Axial STIR MR image showing hyperintense irregularities in the right submandibular gland (arrow) and the adjacent fat tissue. (B) Right submandibular gland at autopsy: hemorrhage involving the lateral and distal parts of the gland was apparent (arrows). (C) Histological section (r 25) showing traumatic interlobular hemorrhage (arrows).
© 2009 by Taylor & Francis Group, LLC
338
The Virtopsy Approach
10. Poquet, E., A. Dibiane, C. Jourdain, M. el Amine, A. Jacob, and M. N. Escure. 1995. Blunt injury of the larynx by hanging. X-ray computed tomographic aspect. J Radiol 76:107–09. 11. Reay, D. T., W. Cohen, and S. Ames. 1994. Injuries produced by judicial hanging. A case report. Am J Forensic Med Pathol 15:183–86. 12. Thali, M. J., K. Yen, W. Schweitzer, et al. 2003. Virtopsy, a new imaging horizon in forensic pathology: virtual autopsy by postmortem multislice computed tomography (MSCT) and magnetic resonance imaging (MRI)—a feasibility study. J Forensic Sci 48:386–403. 13. Wallace, S. K., W. A. Cohen, E. J. Stern, and D. T. Reay. 1994. Judicial hanging: postmortem radiographic, CT, and MR imaging features with autopsy confirmation. Radiology 193:263–67. 14. Yen, K., M. J. Thali, E. Aghayev, et al. 2005. Strangulation signs: initial correlation of MRI, MSCT, and forensic neck findings. J Magn Reson Imaging 22:501–10. 15. Aghayev, E., C. Jackowski, M. Sonnenschein, M. J. Thali, K. Yen, and R. Dirnhofer. 2006. Virtopsy hemorrhage of the posterior cricoarytenoid muscle by blunt force to the neck in postmortem multislice computed tomography and magnetic resonance imaging. Am J Forensic Med Pathol 27:25–29. 16. Jankovich, L. and J. Incze. 1933. Hemorrhage of the neck lymph nodes in death by hanging. Dtsch Z Gerichtl Med 20:122–33.
D3.8 DROWNING: POSTMORTEM IMAGING FINDINGS Andreas Christe, Peter Vock, and Michael J. Thali
D3.8.1 INTRODUCTION A generally used definition of drowning is given by Roll [1]: “Death by drowning is the result of a hampering of respiration by obstruction of mouth and nose by a fluid medium (usually water).” Generally bodies retrieved from water may have (1) died of natural disease before falling into the water, (2) died of natural disease while already in the water, (3) died from injury before being thrown into the water, (4) died of injury while in the water, (5) died of effects of immersion other than drowning, or (6) died of drowning. According to the World Health Organization the annual worldwide incidence of death by drowning is about 400,000 [2]. Death from drowning is more common in young children [3]. In adults suicide is a frequent cause of drowning and is often associated with past psychiatric history [4]. Drowning has especially been related to young adults who are under the influence of alcohol and other drugs while being near water [5,6]. The pathophysiological aspects of drowning are complex, and controversial opinions coexist about typical and atypical drowning. We prefer the differentiation between dry and wet drowning. Dry drowning, which occurs in 10% of cases, happens without aspiration of fluid in the lung [7,8] and occurs due to a laryngospasm because of hypoxia and stimulus of water in the upper respiratory ways (larynx) [9,10]. The
© 2009 by Taylor & Francis Group, LLC
remaining 90% of drowning cases are considered wet drowning: The volume and composition of aspirated fluid determine the physiologic basis of hypoxemia [11–13]. Aspiration of fresh water dilutes the pulmonary surfactant and makes alveoli unstable [14], which leads to the collapse and atelectasis of some of the alveoli. Intrapulmonary shunt occurs with considerable pulmonary venous admixture [15]. The appearance of hypotonic fresh water in the alveoli is not usually a problem, because it is rapidly absorbed into the pulmonary and systemic circulation with dilution of the blood and hypervolemia [16]. The passive postmortal inflow of water is low [17–19]. On the other hand, aspiration of sea water leads to perfused and fluid-filled alveoli. The hypertonic sea water pulls additional fluid from the plasma, leading to pulmonary edema and hypovolemia and adding to the ventilation-perfusion abnormality [13,20]. Intrabronchial water can cause to substantial bronchospasm, which results in pulmonary emphysema, also known as emphysema aquosum [21]. The fluid aspiration followed by bronchospasm is more multifocally scattered throughout the lung than diffuse, which results in a mosaic pattern of dry, hypoperfused and wet, hyperperfused lung areas [21]. Pulmonary arterial hypertension and hypervolemia lead also to heart failure, accentuated on the right side. During the act of conscious drowning, a lot of water is swallowed [22], which leads to a distention of the stomach and duodenum. In addition, an inflow of water into the paranasal sinuses may frequently occur [23].
D3.8.2 TYPICAL SIGNS OF DROWNING IN POSTMORTEM IMAGING The typical signs of drawning in autopsy [24] can also be seen in postmortem imaging. D3.8.2.1 Aspiration Almost every case of drowning shows some content in the airways both in multislice computed tomography (MSCT) and in magnetic resonance imaging (MRI). Most of the fluid is found in the main bronchi (Figure D3.8.1), where on average two thirds of the luminal volume is filled with liquid. But also half of the volume of trachea and small bronchi are filled by liquid at MSCT. A lot of the cases show also a small amount of watery content in the pharynx at MSCT. Generally, postmortem fluid in the airways or the lung is common also in nondrowning cases, but only a small volume is present: on average, 10% to 20% of the volume of the bronchi is filled with liquid and 10% show minimal tracheal aspiration. Aspiration into the trachea and the main bronchi is significantly more severe in drowning (Table D3.8.1) and typically shows high-attenuation sediment in the airways [25]. Any multifocal, ill-defined air-space densities in dependent lung areas with normal calibers of the vessels can be considered as signs of aspiration into the lung. Radiologically, 60% of the drowning cases aspirate fluid into the lung (Figure D3.8.1).
Incident-Specific Cases
A
339
diameter) can be measured on axial images as an indirect sign of heart failure. Interestingly, the CTR does not show a significant difference between drowning cases (0.47) and nondrowning cases (0.48). In the analysis of individual size of the heart chambers, only the right ventricle is slightly enlarged compared with nondrowning cases (Table D3.8.1), probably due to pulmorary arterial hypertension. D3.8.2.4 Pulmonary Edema
B
Diffuse or multifocal air space consolidation with interstitial opacities in dependent lung areas surrounding enlarged pulmonary vessels with accompanying pleural effusion can be considered as signs of pulmonary edema. Of the drowning cases, 50% show pulmonary edema at MSCT or MRI, (Figure D3.8.3) ranging from interstitial to alveolar transsudation. The average weight of both lungs is about 1,000 g at autopsy in drowning cases (range 600–1600 g). Pleural effusion can be found in 70% (80% at autopsy). The typical plume of froth is detected at MSCT by measuring the density of the bubbles in the fluid of the airways, as gas has a density of minus 1000 Hounsfield units (HU). But in postmortem imaging and in autopsy only 20% of the drowning victims have a plume of froth (Figure D3.8.4). D3.8.2.5 Hemodilution
FIGURE D3.8.1 Central and peripheral fresh water aspiration in MSCT. (A) Note the fluid in the main bronchi (white arrows) with an air/fluid level (arrow head) on the left side. Air space consolidation in the upper lobes represents aspiration in the lungs (asterisks). (B) The lung-window at a lower level presents coarse nodular lung densities in the middle lobe, compatible with aspiration (asterisk). In the lower lobes there is a patchy mosaic pattern of hypoperfused (hypodense) and hyperperfused (hyperdense) lung areas. Note the diameter of the vessels: In the hypodense areas the vessels appear smaller (arrow) than in the edematous hyperdense areas (arrow head).
D3.8.2.2 Emphysema Aquosum For the assessment of emphysema aquosum, the stance of the dome of the right diaphragm on the basis of the anterior ribs should be determined. In drowning cases, the dome of the diaphragm is located at an average level of the fifth anterior rib on the right side (rib range 4–6). The position of the diaphragm is generally higher at the level of the fourth anterior right rib (rib range 4–6) (Table D3.8.1). The position of the diaphragm can be confirmed at autopsy. Additionally, the bronchial diameters compared with the diameters of the accompanying pulmonary artery in MSCT can be used as a sign of bronchial constriction, hypothesizing that a ratio of 0.9 (broncho-arterial coefficient) might reflect bronchospasm. A consecutive mosaic pattern of hypoperfused and hyperperfused lung areas is present in 60% of drowning cases in both autopsy and MSCT (Figure D3.8.2).
Since it is known that people with anemia have a lower density of blood [26]—because of the low level of hemoglobine, which is a major contributor to the density of blood in MSCT—hemodilution of blood in each case can be assessed. It is rather difficult to measure the accurate density of blood after sedimentation, but measurement in the right atrium with a large region of interest (ROI) shows significant lower densities in cases of drowning (Figure D3.8.5). In drowning cases the average blood density in the right atrium is 50 HU, where nondrowning cases show a significantly higher mean density of 64 HU (Table D3.8.1). D3.8.2.6 Distension of Stomach and Duodenum The average volume of the stomach in MSCT of drowning victims is about 500 ml, ranging from 50 to 1,200 ml. Interestingly, nondrowning cases show only about half of this volume (Table D3.8.1, Figure D3.8.6). The content of the stomach in drowning cases is diluted by water and measures an average of 20 HU in MSCT; generally the content measures about 40 HU (Table D3.8.1). In 90% of the drowning cases a distension of the duodenum by watery content is additionally present. But 60% of nondrowning cases also demonstrate a duodenal distension (Table D3.8.1).
D3.8.2.3 Heart Failure
D3.8.2.7 Water in the Paranasal Sinuses
To assess the size of the heart the cardio-thoracic ratio (CTR maximum transverse cardiac diameter/transverse thoracic
Every case of drowning presents with fluid in the paranasal sinuses, especially in the maxillary (Figure D3.8.7) and
© 2009 by Taylor & Francis Group, LLC
340
The Virtopsy Approach
TABLE 3.8.1 Signs of Drowning in Order of Stignificance Compared to Non-Drowning Cases [24]
Signs of Drowning in Order of Significance (Wilcoxon-test) in Drowning vs. Non-Drowning Cases Aspiration into main bronchi* Water in sphenoidal sinus* Emphysema (rib level of diaphragm dome) Tracheal aspiration* Water in maxillary sinus* Bronchospasm (bronchial-arterial coefficient) Right ventricle size (semiquantitative from 0 to 3) Water in frontal sinus* Density of airways contents (HU**) Hemodilution (HU** of blood) Density of stomach content (HU**) Duodenal size (semiquantitative from 0 to 3) Aspiration into peripheral bronchi* Laryngeal spasm (glottic distance in cm) Stomach volume (ml) Heart size (semiquantitative from 0 to 3) Pharyngeal aspiration* Aspiration into the lung* Pulmonary edema* Size of left atrium (semiquantitative from 0 to 3) Size of inferior caval vein (cm^2) Water in ethmoidal sinus* Water in mastoid cells* Density of pleural effusion (HU**) Size of left ventricle (semiquantitative from 0 to 3) Distance between lungs (anterior mediastinum in cm) Size of right atrium (semiquantitative from 0 to 3) CTR (cardio-thoracic-ratio) *
amount, semiquantitative from 0 to 3
**
Hounsfield Units
Significance*** 3 3 3 3 3 3 2 2 2 2 2 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
Mean Drowning (n=10) 1.90 2.00 5.00 1.20 1.40 0.84 0.40 1.10 21.90 50.00 20.30 1.20 1.30 0.60 470.00 0.30 0.80 0.70 0.70 0.10 4.72 0.80 0.40 21.00 0.10 1.47 0.70 0.47
sd+/0.57 0.94 0.67 1.03 0.52 0.12 0.52 0.99 10.91 16.17 13.74 0.63 0.67 0.16 323.35 0.48 0.42 0.67 0.82 0.32 2.31 0.42 0.52 20.08 0.32 2.48 0.67 0.09
Mean NonDrowning (n=20) 0.37 0.37 4.00 0.11 0.58 1.04 0.00 0.21 40.45 64.05 38.95 0.63 0.79 0.48 275.50 0.05 0.47 0.37 0.33 0.00 5.53 0.58 0.26 30.00 0.05 1.54 0.68 0.48
sd+/0.60 0.50 0.53 0.32 0.51 0.14 0.00 0.42 16.24 9.74 17.99 0.50 0.63 0.14 222.51 0.23 0.51 0.50 0.49 0.00 1.34 0.51 0.45 10.00 0.23 1.67 0.48 0.05
*** 3: p<0.001; 2: p<0.01; 1: p<=0.05; 0: p>0.05
sphenoidal sinuses. Frontal sinuses have a watery content in 70% and ethmoidal sinuses in 80% of drowning cases.
D3.8.3 CONCLUSION One of the classic findings of drowning at autopsy is Paltau’s spots; these are pleural hemoglobin deposits from hemolysis. Autopsy reveals this finding in 70% of the drowning cases, but the spots are much too small to be visualized by MSCT. It might be possible that the susceptibility artifacts of hemoglobin in MRI might uncover Paltau’s spots in special sequences. Similar experience was made in brain hemorrhage: With fast low-angle shot (FLASH) MRI sequences it is possible to detect tiny amounts of blood [27] that cannot be seen in normal T1- and T2-weighted sequences of the brain. But further studies on this issue are necessary. Otherwise
© 2009 by Taylor & Francis Group, LLC
all the typical forensic signs of drowning can be seen in MSCT or MRI, such as aspiration, emphysema aquosum, mosaic pulmonary edema, and distented stomach and duodenum. Additionally, we could document and directly measure bronchospasm in MSCT, water in the paranasal sinuses, and hemodilution in MSCT, which is rather complicated or impossible with classical autopsy. The pathophysiologic hypoperfused and hyperperfused areas in the lung (mosaic pattern) with narrow or dilated pulmonary arteries are much better visualized in the axial images of the MSCT than during classical autopsy. The bronchial-arterial coefficient is significantly decreased in drowning cases and is a strong sign of bronchospasm: This coefficient decreases due to a small diameter of the bronchus (bronchospasm) or due to a large arterial diameter, like in reactively hyperperfused lung areas.
Incident-Specific Cases A
341 C
B
FIGURE D3.8.2 Mosaic pattern of lung perfusion in aspiration. (A) Hyperperfused lung areas with ground-glass appearance (asterisks) and dilated pulmonary arteries (arrows) are found in contrast to the hypoperfused lung parenchyma with narrow arteries (arrow head). (B) Autopsy correlation of the mosaic perfusion with hyperperfused, dark red lung areas and hypoperfused overinflated pale areas. (C) Hyperperfused upper lobe with hyperdense lung parenchyma (asterisk). Note the increased diameter of the pulmonary artery (arrow) compared to the accompanying bronchus (arrow head).
A
B
B
C
FIGURE D3.8.4 Plume of froth. Sagittal reconstruction of the head at soft tissue window (A) and autopsy correlation (B, C). Froth in the nostrils (arrow), upper airways, nasal cavity, pharynx and trachea (arrow head) (C).
A
B
FIGURE D3.8.5 Hemodilution. (A) Measurement of the blood density in the right dilated atrium in a drowning case shows 45 HU, indicating possibly increased water content (arrow). Note the additional gas in the right ventricle. (B) A case from the control group presents a “normal,” postmortal density of the blood in the right atrium of 70 HU (arrow).
A
FIGURE D3.8.3 Pulmonary edema in MRI. Coronal T2-weighted fat-saturated MRI-sequence shows hyperintense multifocal air space consolidation throughout the lung (white asterisks) with accompanying pleural effusion (arrow). Additional signs of drowning are aspirated fluid in trachea and main bronchi (arrow head) and swallowed water in the distended stomach (black asterisk).
© 2009 by Taylor & Francis Group, LLC
B
FIGURE D3.8.6 Distention of stomach. (A) Swallowed water in a distended stomach with an air fluid level (arrowhead) and alimentary leavings swimming close to the top (arrow). Note the hypodense liver indicating fatty liver as an additional finding (31HU). (B) Autopsy image of the stomach fluid content.
342
The Virtopsy Approach
allows for a sampling of probes from body tissues and fluids for subsequent examinations such as the evidence of diatoms. However, CT might be used to get samples from the most important sites nearly noninvasively. The first results of the use of postmortem CT-guided biopsy have been recently reported [28], and the addition of CT-guided biopsy might increase the role of postmortem CT in forensic medicine. On the other hand, MSCT and MRI provide an objective, compact, and distributable documentation of the full body in a noninvasive way [29,30].
D3.8.4 REFERENCES
FIGURE D3.8.7 Water in the paranasal sinuses on an axial CT image. The maxillary sinuses show air/fluid levels in a drowning victim (arrowhead). The sphenoidal sinus is totally filled with fluid (arrow).
No standard values exist for postmortem blood densities (HU), but significant lower densities can be measured in cases of drowning. This level needs to be confirmed in a larger study. The size of the right atrium, like the size of the inferior caval vein, although large does not seem to be specific for drowning, as nondrowning cases show similar data. This is possibly due to the general blood congestion during death. The amount of water in the lung is rather small because of the quick hemodilution effect of fresh water in the lung. A secondary mosaic pulmonary edema due to bronchospasm with consecutive hyperperfused lung areas is much more specific. The stomach and duodenum are distended, but interestingly only the duodenal size of the drowning victims shows sa significant difference compared with the nondrowning cases (Table D3.8.1). As the aspirated and swallowed fluid in drowning is fresh water, a low density in the stomach is expected despite the admixture with the rather dense contents of the stomach. So the content shows an average density of 20 HU, and the normal content of nondrowning cases is 39 HU. Although the described findings of drowning are not specific, those combinations such as froth in the airways, emphysema aquosum, and mosaic patterns of the lung in a victim retrieved from fresh water can serve as highly suspicious for drowning. All our cases were exclusively freshwater drowning victims. Since the pathophysiology of drowning in salt water is different, the radiological appearance of the examined signs might deviate from our data. The high osmolality of aspirated salt water is expected to pull water into the lung, leading to substantial lung edema and hypovolemia. By comparing postmortem CT and autopsy as examination methods to detect the cause of death, autopsy easily
© 2009 by Taylor & Francis Group, LLC
1. Roll HF. Leerboek der Gerechtelijke Genesekunde voor de scholen tot opleiding van Ind. Artsen’s—Gravenhage. Martinus Nijhoff, 1918. 2. World Health Organization (WHO). Reducing risks, promoting health life. The world health report; Geneva: WHO, 2002. 3. Web-Based Injury Statistics Query and Reporting System (WISQARS). Leading causes of death reports. National center for injury prevention and control and national, 2000. 4. Byard RW, Houldsworth G, James RA, and Gilbert JD. Characteristic features of suicidal drownings: a 20-year study. Am J Forensic Med Pathol 2001;22(2):134–38. 5. Mackie IJ. Patterns of drowning in Australia, 1992–1997. Med J Aust 1999;171(11–12):587–90. 6. Cummings P and Quan L. Trends in unintentional drowning: the role of alcohol and medical care. JAMA 1999;281(23):2198–202. 7. Modell JH, Graves SA, and Ketover A. Clinical course of 91 consecutive near-drowning victims. Chest 1976;70(2):231–38. 8. Kringsholm B, Filskov A, and Kock K. Autopsied cases of drowning in Denmark 1987–1989. Forensic Sci Int 1991;52(1):85–92. 9. Modell JH. Drowning. N Engl J Med 1993;328(4):253–56. 10. Olshaker JS. Near drowning. Emerg Med Clin North Am 1992;10(2):339–50. 11. Modell JH and Moya F. Effects of volume of aspirated fluid during chlorinated fresh water drowning. Anesthesiology 1966;27(5):662–72. 12. Modell JH, Gaub M, Moya F, Vestal B, and Swarz H. Physiologic effects of near drowning with chlorinated fresh water, distilled water and isotonic saline. Anesthesiology 1966;27(1):33–41. 13. Modell JH, Moya F, Newby EJ, Ruiz BC, and Showers AV. The effects of fluid volume in seawater drowning. Ann Intern Med 1967;67(1):68–80. 14. Giammona ST and Modell JH. Drowning by total immersion. Effects on pulmonary surfactant of distilled water, isotonic saline, and sea water. Am J Dis Child 1967;114(6): 612–16. 15. Modell JH, Moya F, Williams HD, and Weibley TC. Changes in blood gases and A-aDO2 during near-drowning. Anesthesiology 1968;29(3):456–65. 16. Tabeling BB and Modell JH. Fluid administration increases oxygen delivery during continuous positive pressure ventilation after freshwater near-drowning. Crit Care Med 1983;11(9):693–96. 17. Fuller RH. The clinical pathology of human near-drowning. Proc R Soc Med 1963;56:33–38.
Incident-Specific Cases
18. Fuller RH. The 1962 Wellcome prize essay. Drowning and the postimmersion syndrome. A clinicopathologic study. Mil Med 1963;128:22–36. 19. Modell JH and Davis JH. Electrolyte changes in human drowning victims. Anesthesiology 1969;30(4):414–20. 20. Modell JH. The pathophysiology and treatment of drowning. Acta Anaesthesiol Scand Suppl 1968;29:263–79. 21. Aghayev E, Thali MJ, Sonnenschein M, et al. Fatal steamer accident; blunt force injuries and drowning in postmortem MSCT and MRI. Forensic Sci Int 2005;152(1): 65–71. 22. de Boer J, Biewenga TJ, Kuipers HA, and den Otter G. The effects of aspirated and swallowed water in drowning: Sea-water and fresh-water experiments on rats and dogs. Anesthesiology 1970;32(1):51–59. 23. Hotmar P. Nachweis von Flüssigkeit in den Nasennebenhölen als mögliches diagnostisches Zeichen des Ertrinkungstodes. Arch Kriminol 1996;198. 24. Christe A, Aghayev E, Jackowski C, Thali MJ, and Vock P. Drowning—postmortem imaging findings by computed tomography. Eur Radiol 2007;18(2):283–90. 25. Levy AD, Harcke HT, Getz JM, et al. Virtual autopsy: twoand three-dimensional multidetector CT findings in drowning with autopsy comparison. Radiology 2007;243(3): 862–68. 26. Title RS, Harper K, Nelson E, Evans T, and Tello R. Observer performance in assessing anemia on thoracic CT. Am J Roentgenol 2005;185(5):1240–44. 27. Hartmann M, Jansen O, Deinsberger W, Vogel J, and Sartor K. MRI of acute experimental intracerebral hematoma. Neurol Res 2000;22(5):512–16. 28. Aghayev E, Thali M, Sonnenschein M, Jackowski C, Dirnhofer R, Vock P. Post-mortem tissue sampling using computed tomography guidance. Forensic Sci Int 2007; 166(2–3):199–203. 29. Thali MJ, Braun M, Buck U, et al. VIRTOPSY— scientific documentation, reconstruction and animation in forensic: individual and real 3D data based geo-metric approach including optical body/object surface and radiological CT/MRI scanning. J Forensic Sci 2005;50(2): 428–42. 30. Thali MJ, Yen K, Schweitzer W, et al. Virtopsy, a new imaging horizon in forensic pathology: virtual autopsy by postmortem multislice computed tomography (MSCT) and magnetic resonance imaging (MRI)—a feasibility study. J Forensic Sci 2003;48(2):386–403.
343
TABLE D3.9.1.1 Forensic Questions in Charred Bodies Identification Vital reactions Toxicological analysis Cause and manner of death
application of heat. Human tissues can survive only within a relatively narrow range of temperatures, approximately 20nC to 44nC. When external heat is applied, the extent of damage depends on the following (Table D3.9.1.2): r The applied temperature r The ability of the body surface to conduct excess heat away r The duration of the heat applied The relationship between temperature and time is important, for it is sometimes forgotten that relatively low temperatures, (even as low as 44nC), can cause tissue damage when sustained too long [2,6]. In forensic medicine it is more often seen when a disabled or older person becomes unconscious (e.g as due to syncope), within close range of a fire or heating place, like the oven, have an average 9.7% of the body surface burned compared to 8.4% in healthy or young people, with no loss of consciousness [7]. D3.9.1.2 Stages of Burn Wounds Wounds from burning are classified in three or four stages. The three-stage classification contains the following stages [2,4,5]: r First degree: superficial burns r Second degree: partial thickness burns r Third degree: full thickness burns The four-stage classification contains these stages:
D3.9 THERMAL DAMAGE D3.9.1 HEAT AND BURNS Danny Spendlove and Michael J. Thali
r r r r
First degree: superficial, epidermal burn wounds Second degree: superficial, deep dermal wounds Third degree: transdermal thermal wounds Fourth degree: subcutaneous burning wounds
D3.9.1.1 Introduction Damage to the tissues arising from the application of heat is commonly encountered in forensic pathology and can provide a challenging forensic problem in examination, interpretation, and conclusion to with regard identification, vital reactions, toxicological analysis, and determination of the cause and manner of death [1–5] (Table D3.9.1.1). Heat injury may arise following a defect in body temperature control or, more, commonly, from the external
© 2009 by Taylor & Francis Group, LLC
TABLE D3.9.1.2 Extent of Damage Extent of Damage Depends on the Following Applied temperature Ability of the body surface to conduct away the excess heat Duration of the heat applied
344
The Virtopsy Approach
TABLE D3.9.1.3 First-Degree Burn Wounds
TABLE D3.9.1.5 Third-Degree Burn Wounds
First-Degree Burn Wounds or Superficial Wounds
Third-Degree Burns Can Have Different Appearances
Reddish skin color, which is painful Skin heals within 8–10 days
Can be either black or white Depending on whether the skin was baked, as with a fire, or boiled, as with hot water Skin nerves are destroyed and no pain is felt by the patient
First-degree burning wounds or superficial wounds (Table D3.9.1.3) are recognizable due to their reddish skin color, which are painful. These can be compared with Sunburn. Normally these wounds heal within 8–10 days. In corpses, the erythema disappears because of the absence of circulation. Second-degree burns (Table D3.9.1.4) are characterized by blisters that are surrounded by first-degree wounds. The skin can be moist and is painful. The deep second-degree wounds go through the basal stratum of the skin, with the skin less moist than the superficial ones. Normal healing time takes about 2–3 weeks and causes a scar; the size of the scar depends on the skin characteristics and exact depth of the burn. Third-degree burns (Table D3.9.1.5) can have different appearances; they can be either black, when the skin was baked, e.g., with a fire, or white, like the skin was cooked with hot water. Because of the depth of the wound (past the subcutis), the skin nerves are destroyed, and no pain is felt by the burn victim. Although spontaneous healing is possible, it takes quite a while and can involve some inflammatory reactions. A better result is given by a transplantation of the skin (full thickness graft). Fourth-degree wounds with charring reach deeper in the soft-tissue layers. Sometimes it is difficult to differentiate between second- and third-degree burns, especially when the blisters are destroyed. In the living, sensibility can be tested by a sharp object, like a needle, which causes pain in second-degree burns. Of course, it is possible to document the lesions with the previously described photogrammetry and 3D surface scanning for later reconstructions (see Chapter B1, “External Body Documentation”).
parts into portions of 9% each. Thus, the arms are each 9%; the front and the back of each leg are 9% (so each leg is 18%); the body front and back are each 18%; the head is 9%; and the genitals count as 1%. With children, a variation on the rule of nines is used: The head is counted as 18%; the legs are each 14%; the front and the back of the torso are each 18%; and the arms count for 9% each. To make a gross evaluation of the burned skin, the palm of the hand of the patient can be used and counts for about 1% (Figure D3.9.1.1a) [1,4]. The area of involvement is used to make survival estimation. For example, having more than 30% deep burns in humans has a serious prognosis, and with burns of 50% or more in the body, a bad prognosis is given [4]. The exact prognosis depends on the depth of the burns and the area of involvement. The burning index is used for the clinical assessment of survival chance. Add the age of the patient and the area of involvement of the second- and third-degree wounds: An index of 80 or less equates to little risk of life; 80–120 is
LAO/RAO CRAN/CAUD
0 3
D3.9.1.3 Area of Involvement A “Rule of Nines” chart is used for measuring the extent and severity of burns. The rule divides the human adult body
TABLE D3.9.1.4 Second-Degree Burn Wounds Second-Degree Wounds Blisters, surrounded by first-degree wounds Skin can be moist and is painful Deep Second-Degree Wounds Through the stratum basal of the skin Skin less moist than superficial wounds Healing time takes about 2–3 weeks and causes scarring
© 2009 by Taylor & Francis Group, LLC
FIGURE D3.9.1.1A CT 3D volume-rendering technique (VRT) reconstruction of the surface of the corpse. Visible is the massive destruction caused by fire, which has affected the muscles and the soft tissue.
Incident-Specific Cases
TABLE D3.9.1.6 Postmortem Heat Injury
345
LAO/RAO CRAN/CAUD
1 5 4
Causes of Postmortem Heat Injury Evaporation Heat-dependent shrinking Direct flame influence
acute risk; and over 120 indicates that survival is unlikely. With inhalation trauma, survival becomes worse. As described before, the hand is used for estimation of the involved area, although the exact area can be measured by photogrammetry and 3D surface scanning documentation (see Chapter B1). This method gives a clear image of the involved burned areas and makes reconstructions possible. D3.9.1.4 Postmortem Heat Injury Postmortem heat injury is caused by evaporation, heat-dependent shrinking, and direct-flame influence (Table D3.9.1.6). Direct fire, internal steam generation (as a result of cooking organs), and the change of the fluids constellation in the burned body (Figure D3.9.1.1b) cause destruction in a human corpse. The internal organs become cooked, the organs shrink (so-called puppet organs) (Figure D3.9.1.2), and the body becomes blown up to a certain degree, just like in decomposition, up to a certain degree. After a certain time, the limit of body swelling and evaporation is reached; therefore, the skin ruptures due to tightening and, for example, the intestines are forced out of the abdomen. The skin at this point usually is a leathery, hard consistency. Because of the shrinking skin, lesions become smaller, change in shape, and may wander to the center of the heat source (Figure D3.9.1.3a, Figure D3.9.1.3b, Figure D3.9.1.3c, and Figure D3.9.1.3d) [1,2,4,5]. In this state, the mouth is usually opened with shrunken lips. The lips protect the teeth for a long period before heat can induce changes in them. The eyes are normally closed, and the lids can only be opened with difficulty and incompletely with force. Protrusion of the tongue is caused by heatrelated shrinkage of the soft tissue of the neck. The shrinkage of the soft tissue of the neck can also be responsible for petechial hemorrhages in the eyelids and conjunctivae when circulation is still present. These hemorrhages can be seen as a vital reaction (see also Chapter D2.3, “Vital Reactions and Vital Signs”). The epidural hematoma is a frequent finding in heat injuries (Figures D3.9.1.4 A–F) [8]. This postmortem effect is a result from fluid shifting from the diploe and the venous sinuses to the room between the dura and skull and is usually found under the site of charring of the skull. The evaporation of brain fluid can cause the skull to rupture (Figure D3.9.1.4A, Figure D3.9.1.4C, and Figure D3.9.1.4D). The typical posture of charred bodies is a pugilistic, or boxer-like, attitude and is caused by the loss of fluids and
© 2009 by Taylor & Francis Group, LLC
FIGURE D3.9.1.1B 3D CT reconstruction, coronal view on the internal organs. Heatshrinked, puppet organs can be seen.
the shrinkage of the muscles, tendons, and sinews, and is characterized by abduction in the shoulder and hip joints and flexion in the elbow, hip, and knee joints. The stronger flexor muscles are responsible for this position, which is strongly seen in the hands and wrists and causes fists (Table D3.9.1.7, Figure D3.9.1.5).
LAO/RAO CRAN/CAUD
FIGURE D3.9.1.2 Sagittal view in a CT 3D reconstruction.
56 –3
346
FIGURE D3.9.1.3A 20-year-old male who was drunk and fell asleep with a burning cigarette, after which time the blankets began to burn and later the house burned as well. Death due to carbon monoxide intoxication. Sagittal view with an osteo windowing. Remarkably the skin destruction is clearly visible.
D3.9.1.5 Vital Reactions† The inhalation of hot gases causes damage to the mucosa of the respiratory tract, which is not seen in a body that is dead before being exposed to fire. Edema, mucosal bleeding, and vasicular detachment may be indicative of hot gas inhalation and can be seen as a vital reaction (Table D3.9.1.8.) When respiration is still intact, the inhalation of soot is also seen as a vital sign (Figure D3.9.1.6), but it has to be found in the deeper respiratory tract and comes along with great amounts of slime. Parallel to this, soot may be swallowed and be found with examination of the stomach content. Another vital reaction is—as already mentioned—petechial hemorrhages that are caused by intact circulation and shrinkage of the soft tissue of the neck and upper thorax, but these can also be caused by strangulation while still alive. High levels of carbon monoxide and carbon cyanide which comes free as a result of incomplete burning, are classically seen as a vital sign, because of the inhalation of the gas, the color of the internal organs changes to bright red. Notice that levels of carbon monoxide up to 10% can be found in smokers. Crow’s feet are also considered to be a vital sign in burning, because of the squinting of the eyes as a reflex to the smoke or flashover. Another explanation for these crow’s feet is the shrinkage of the skin due to heat, resulting in smoothing of the wrinkles of the face, which results in the visibility of the unsooted base of wrinkles [1–5] (Figure D3.9.1.8a, Figure D3.9.1.8b, and Figure D3.9.1.8c). †
See also Chapter D2.3, “Vital Reactions and Vital Signs”
© 2009 by Taylor & Francis Group, LLC
The Virtopsy Approach
FIGURE D3.9.1.3B Coronal view (osteo windowing). Puppet organs.
D3.9.1.6 Forensic Examination The correct forensic examination is expected to give answers to questions in respect to the following [1,8]: r Identification r Vital reactions indicating that the decedent had been alive when the fire started r The toxicological examination r Cause of death and injuries r Manner of death
FIGURE D3.9.1.3C Axial view (CT image). Destruction of the skin.
Incident-Specific Cases
347
FIGURE D3.9.1.3D Axial view of a CT image of the same person, through the pelvis.
FIGURE D3.9.1.4B MR sagittal view of the epidural heat hematoma (yellow cross).
The external surface of the body is often greatly destroyed, and visual identification, due to the damage caused by intensive heat and flame, is frequently impossible. Regardless of the severity of the superficial destruction, though, the teeth and organs of a body recovered from a fire are usually well preserved. Teeth and dental work are remarkably resistant
to fire and can be used for a dental identification (see also Chapter D1, “Radiologic Identification”) [9–11]. In addition, tissue and body fluids are usually available for DNA and comprehensive toxicological analysis of, for example, carbon monoxide (CO) and hydrogen cyanide from the burning and to differentiate between swallowed ethanol and other alcohols or substances created by the decomposition process.
LAO/RAO CRAN/CAUD
9 18
FIGURE D3.9.1.4A CT 3D VRT reconstruction of a male who died in a house fire. Visible is the traumatic opening of the skull and the epidural fire hematoma, as well as the soft-tissue destruction.
© 2009 by Taylor & Francis Group, LLC
FIGURE D3.9.1.4C Sagittal CT multiplanar reconstruction (MPR) of this person. Epidural fire hematoma (blue cross). Traumatic fracture of the skull with splitting of the tabulae (green cross).
348
The Virtopsy Approach
FIGURE D3.9.1.4F Epidural heat hematoma in autopsy.
FIGURE D3.9.1.4D Axial CT view of the fractured skull. The skull fractures are clearly seen. Air is present in the posterior part of the skull (black areas). The epidural fire hematoma is visible (red cross).
Determining whether the injuries that resulted from heat of direct flame contact were caused postmortem or while the victim was still alive requires great knowledge and experience on the part of the forensic pathologist. The organs, the injury pattern, and the so-called vital signs and vital reactions have
to be studied and combined with the toxicological results in order to finally indicate the cause of death. The scene and criminal investigations based on the cause of death define the manner of death. Following an old forensic rule, x-ray analysis of the entire burned body serves in locating foreign bodies or fractures [3,8,12,13]. The application of radiological methods for answering forensic questions dates back to the beginnings of x-ray diagnosis [12] (Figure D3.9.1.9A, Figure D3.9.1.9B, Figure D3.9.1.9C, Figure D3.9.1.9D, and Figure D3.9.1.9E). D3.9.1.7 Identification The desiccation of skin and tissues, skeletal fractures, and pulverization of the vertebral disks by the effects of heat completely alter the appearance and the length of the body [3]. Visual or fingerprint identification is often not possible
TABLE D3.9.1.7 Findings Organ Shrinkage (So-Called Puppet Organs) Body becomes blows up till a certain degree Skin ruptures due to tightening The skin is usually a leathery, hard consistency Skin Shrinkage Lesions become smaller, change in shape, and may wander to the center of heat source Mouth opened with shrunken lips Eyelids closed Soft-Tissue Shrinkage of the Neck
FIGURE D3.9.1.4E MR axial view of the same patient. The epidural fire hematoma is very visible (yellow cross).
© 2009 by Taylor & Francis Group, LLC
Tongue protusion Petechial hemorrhages in eyelids and conjuntivae Pugilistic or boxer-like attitude Epidural hematoma
Incident-Specific Cases
349
FIGURE D3.9.1.6 Soot aspiration after a car caught on fire. Notice the soot in the bronchus.
FIGURE D3.9.1.5 CT 3D VRT reconstruction of a 57-year-old man who caught on fire, caused by lightning, and fell about 13 meters from a cabin onto the ground. Visible are the boxer-like attitude and heat-induced fractures of the left forearm (red arrow), as well as a fracture of the iliac crest (yellow arrow).
(Table D3.9.1.9). Teeth, however, are often more resistant to fire than other DNA material. Nevertheless, the conservation and examination of dental material at the scene and then in the morgue are often problematic [6]. For visual examination of the teeth, a mutilating procedure including excision of the upper and lower jaw is performed. Due to metal artifacts within the teeth and jaws after dental work, satisfactory documentation and identification of the victim cannot be achieved by the ordinary axial crosssectional multislice computed tomography (MSCT). A dedicated CT scanning technique called Dentascan [15]—which is tailored to the teeth, their alveolar fixation, and transplant planning—might be combined with metal artifact reduction algorithms [9,10] and might provide a nondestructive documentation of the dental status. Identification of implants, plates, artificial joints, metal staples, and foreign bodies can be performed by current
MSCT screening techniques (see Chapter D1). These objects can be three-dimensionally localized within the body. Other areas of interest for radiological identification such as the sinuses, the mastoids, the sella tursica, and other skeletal variations have been described by Brogdon [12]. These methods of identification can be applied to the charred body when using CT or magnetic resonance imaging (MRI) postmortem data and comparing it to antemortem data. Radiological postmortem data, combined with the appropriate clinical data, often help the specialist to reach anthropologic conclusions with certainty, such as regarding the individual age of a burned or charred body [8,16].
TABLE D3.9.1.8 Vital Reactions Respiratory tract mucosa damage Edema, mucosal bleeding, vasicular detachment Inhalation of soot in the deeper respiratory tract Soot swallowing Facial petechial hemorrhages High levels of carbon monoxide Crow’s feet
FIGURE D3.9.1.7 Crow’s feet.
© 2009 by Taylor & Francis Group, LLC
350
The Virtopsy Approach
FIGURE D3.9.1.8C Skull at autopsy.
D3.9.1.8 Vital Signs during Fire Exposure in CT FIGURE D3.9.1.8A CT 3D VRT reconstruction of the skull. Only the bony parts of the skull are visible.
FIGURE D3.9.1.8B Overview of the skull from the corpse.
© 2009 by Taylor & Francis Group, LLC
First has to be determined whether the victim was alive when exposed to fire, followed by determination of the cause of death. Soot in the airways and the esophagus, as found in autopsy, was not detected by imaging methods, not even in cases of fire in closed rooms where a lot of soot is produced (Table D3.9.1.10). CT can prove vital reactions, such as air embolism to the heart and blood aspiration to the lung. Although gas release may be due to retraction of blood
FIGURE D3.9.1.9A A 79-year-old man collapsed and fell into a burning fire. He was discovered an hour later in the fire. Heat shrinkage of the brain with epidural fire hematoma (green cross) and air from the boiled organs are visible (black areas), as well as damage to the skin.
Incident-Specific Cases
FIGURE D3.9.1.9B Boiled brain due to fire. The air is clearly visible.
351
FIGURE D3.9.1.9D Autopsy overview of the burned skull.
products or formation of coagulum secondary to heat effect, air embolism after traumatic fractures has been observed. The pale color of the organs as a sign of severe hemorrhage cannot be shown by MSCT [17–19]. Blunt trauma or heat can cause fat droplets in the pulmonary blood vessels, causing fat embolism. Fat embolism
FIGURE D3.9.1.9E Fire hematoma at autopsy.
TABLE D3.9.1.9 Identification FIGURE D3.9.1.9C Sagittal view. Epidural fire hematoma (green cross) and boiled brain with air in the tissue (black areas). Note the air on the right side of the neck (red arrow) and the tissue destruction on the left side (yellow arrow).
© 2009 by Taylor & Francis Group, LLC
Visual/fingerprint identification not possible Dental status/skeletal identification Surgical interventions/former injuries DNA
352
The Virtopsy Approach
TABLE D3.9.1.10 Radiological Vital Signs Vital Signs by MSCT and MRI Air embolism Blood aspiration Gastric content aspiration Collapsed aorta (exsanguination) Other Not Yet Visible by MSCT and MRI Fat embolism Soot aspiration Petechial hemorrhages
can be seen as a cause of death in blunt trauma, but it cannot be considered as a vital reaction in death due to burns (Figure D3.9.1.10a, Figure D3.9.1.10b, and Figure D3.9.1.10c) [20–22].
FIGURE D3.9.1.10B Fat embolism, sudan colored, in the lungs.
D3.9.1.9 Alcohol, Drugs, or Other Burn-Specific Analyses (Carbon Monoxide, Hydrogen Cyanide from Burning Polymers)
requested and to use them for highly specific laboratory analyses without cutting the skin and creating external wounds (see Chapter D5, “Biopsy”) [23]. This could help in countries where autopsy is religiously forbidden or refused and therefore hardly ever performed (Figure 3.9.1.11a, Figure 3.9.1.11b, and Figure 3.9.1.11c).
MSCT and MRI cannot show altered chemical components of tissues, because they are primarily morphological methods that show local lesions. Nonetheless, newer MR scanners also perform spectroscopic investigations, showing the relative tissue concentrations of several important chemical compounds [15]. A noncontact, toxicological investigation of tissue and body fluids is conceivable in the future. Another approach might be to gain needle biopsies and fluid samples under image guidance from any tissue or body location
FIGURE D3.9.1.10A Fat embolism in the lungs after blunt trauma (native view, magnification 10 k).
© 2009 by Taylor & Francis Group, LLC
D3.9.1.10 Cause of Death and Injuries Difficulties occur when trying to distinguish fractures caused by physical trauma and fractures due to heat exposure. Heat can cause artificial fractures to the skull. Flame contact with the exterior surface of the skull often causes local defects as a result of incineration of the external table and exposure of the spongy diploic layer of bone [1,3]. Cracks in the skull
FIGURE D3.9.1.10C Fat droplets on the lung tissue in a case of accidental burning (magnification 4x, Sudan stain).
Incident-Specific Cases
FIGURE D3.9.1.11A CT MPR axial view. Because the pressure in the abdomen is rising, a rupture in the abdominal wall can occur and push out the intestines (arrows).
during life are rarely limited to the external table. Often, epidural hematomas are caused by heat, known as heat epidurals, where heat expresses clots of blood and marrow from the skull bone and the clots accumulate between the undersurface of the bone and dura mater [1–5]. The differentiation of an epidural hematoma acquired during lifetime and the
353
FIGURE D3.9.1.11C Overview of the burned body, with the intestines outside the abdominal cavity.
clot-like heat epidural is made by the fact that heat epidural bleedings do not stop at the sutures of the skull, as vital epidural bleedings will stop at the sutures. Splitting of the skin of a burn victim occurs in much the same way as heat fractures of the bones and might be radiologically documented. Imaging methods are even better in localizing and reconstructing the intensity and direction of heat. Other causes of death in burn casualties are as follows [4,5] (Table D3.9.1.11):
r Cyanide intoxication: As already discussed, the altered chemical tissue components cannot be detected by MSCT or MRI. r Flash fire: A flash fire can occur when fire accelerators are used. The flash can cause a spasm within the larynx and the bronchi or a vagal reaction, resulting in apnea. This apnea can be responsible for normal chemical or morphological findings.
TABLE D3.9.1.11 Causes of Death Causes of Death (Not by Imaging Methods)
FIGURE D3.9.1.11B CT MPR sagittal view. Because the pressure in the abdomen is rising, a rupture in the abdominal wall can occur and push out the intestines (arrows).
© 2009 by Taylor & Francis Group, LLC
Carbon monoxide intoxication Cyanide intoxication Flash fire Oxygen deficiency Heatstroke Heat rigor
354
The Virtopsy Approach
r Oxygen deficiency: Every fire needs oxygen to continue to burn. This oxygen consumption can cause an exogenous oxygen depression and can be responsible for deadly suffocation. Imaging methods cannot visualize this cause of death. r Heatstroke: Death due to heatstroke is a shock mechanism that can be caused by a redistribution of the circulated fluids due to heat exposure to the skin. r Heat rigor: Due to heat exposure to the chest, a prompt heat rigor can occur. This can hinder normal functional breathing. Although heat exposure is made visual in the charred body in MSCR and MRI, heat rigor cannot be made visible by imaging methods and is also difficult to prove during autopsy. The determination of the manner of death (e.g., suicide/homicide) was not possible radiologically in all causes of death due to heat exposure. D3.9.1.11 References 1. DiMaio V.J. and D. DiMaio. 2001. Forensic Pathology, 2d ed. CRC Press, London. 2. Saukko P. and B. Knight. 2004. Knight’s Forensic Pathology, 3d ed. Arnold Publishers, Oxford University Press, New York. 3. Spitz W.U. 1993. Medicolegal Investigation of Death. Thomas, Springfield, IL. 4. Brinkmann B. and B. Madea. 2004. Handbuch gerichtliche Medizin [Handbook of forensic medicine, authors’ translation]. Springer Verlag, Berlin. 5. Reimann W. and O. Prokop. 1985. Vademekum Gerichtsmedizin, 4th ed. [Vademecum forensic medicine, authors’ translation]. VEB Verlag Volk und Gesundheit, Berlin. 6. Moritz A.R. and F.C. Henriquez. 1947. Studies of thermal injury: II. The relative importance of time and surface temperature in the causation of curaneous burns. Am J Pathol 23:695–720. 7. Chiu T., P. Pang, S.Y. Ying, and A. Burd. 2004. Syncope and burns. Burns 30:438–42. 8. Thali M.J., K. Yen, T. Plattner, et al. 2002. Charred body: Virtual autopsy with multi-slice computed tomography and magnetic resonance imaging. J Forensic Sci 47(6):1326–31. 9. Jackowski C., E. Aghayev, M. Sonnenschein, R. Dirnhofer, and M.J. Thali. 2006. Maximum intensity projection of cranial computed tomography data for dental identification. Int J Legal Med 120(3):165–67. 10. Jackowski C., A. Lussi, M. Classens, et al. 2006. Extended CT scale overcomes restoration caused streak artefacts for dental identification in CT—3D colour encoded automatic discrimination of dental restorations. J Comput Assist Tomogr 30(3):510–13. 11. Sidler M., C. Jackowski, R. Dirnhofer, P. Vock, and M.J. Thali. 2006. Use of multislice computed tomography in disasters victim identification—advantages and limitations. Forensic Sci Int 169(2–3):118–28.
© 2009 by Taylor & Francis Group, LLC
12. Brogdon B.G. 1998. Forensic Radiology CRC Press, Boca Raton, FL. 13. Thali M.J., K. Yen, W. Schweitzer, et al. 2003. Virtopsy, a new imaging horizon in forensic pathology: virtual autopsy by postmortem multislice computed tomography (MSCT) and magnetic resonance imaging. J Forensic Sci 48(2):386–403. 14. Benthaus S., A. Duchesne, and B. Brinkmann. 1998. A new technique for the post-mortem detection of tooth-coloured dental restorations. Int J Legal Med 111(3):157–59. 15. Kaufman M.J. 2001. Brain Imaging in Substance Abuse: Research, Clinical, and Forensic Applications. Humana Press Inc., Totowa, NJ. 16. Debouit F., N. Telmon, R. Costagliola, P. Otal, F. Joffre, and D. Rougé. 2007. Virtual anthropology and forensic identification: Report of one case. Forensic Sci Int 173(2–3):182–87. 17. Aghayev E., M. Sonnenschein, C. Jackowski, et al. 2006. Postmortem radiology of fatal haemorrhage: measurements of cross-sectional areas of major blood vessels and volumes of aorta and spleen on MDCT and volumes of heart chambers on MRI. AJR 187:209–15. 18. Jackowski C., M.J. Thali, M. Sonnenschein, et al. 2004. Visualization and quantification of air embolism structure by processing postmortem MSCT data. J Forensic Sci 49(6):1339–42. 19. Aghayev E., K. Yen, M. Sonnenschein, et al. 2005. Pneumomediastinum and soft tissue emphysema of the neck in post-mortem CT and MRI; a new vital sign in hanging? Forensic Sci Int 153:181–83. 20. Mudd K.L., A. Hunt, R.C. Matherly, et al. 2002. Analysis of pulmonary fat embolism in blunt force fatalities. J Trauma 48(4):711–15. 21. Inoue H., A. Tsuji, K. Kudo, and N. Ikeda. 2005. Pulmonary fat embolism induced by exposure to high ambient temperature in rats with a fatty liver. Int J Legal Med 119(5):275–79. 22. Taviloglu K. and H. Yanar. 2007. Fat embolism syndrome. Surg Today 37(1):5–8. 23. Aghayev E., M.J. Thali, M. Sonnenschein, C. Jackowski, R. Dirnhofer, and P. Vock. 2007. Post-mortem tissue sampling using computed tomography guidance. Forensic Sci Int 166(2-3):199–203.
D3.9.2 HYPOTHERMIA Lars Oesterhelweg, Emin Aghayev, and Michael J. Thali Hypothermia-related deaths have been reported with a frequency of 0.2 per 100,000 individuals per year in 2002 in the United States [1]. Death from hypothermia not only occurs among those subjected to extreme outdoor temperatures but is also a widespread danger in temperate climates and indoors [2,3]. It is important to know that death from hypothermia is not limited to the cold season but also occurs in the summer. It must be pointed out that no precise degree of Celsius or Fahrenheit can be given that is potentially life-threatening. The vulnerability of individuals to the low temperatures
Incident-Specific Cases
355
TABLE D3.9.2.1 Internal Findings
TABLE D3.9.2.2 External Findings
Spot-like hemorrhages in the gastric mucosa Lung edema Muscular hemorrhages
Purple-red spot-like blotches in nonhypostatic areas Pink livores Paradoxical undressing Hide-and-die syndrome
varies considerably [3,4]. In cases of death from hypothermia in temperate climatic areas, circumstances must be present that lead to a helpless situation of the individual. These situations are mostly based on socioeconomical problems such as advanced age, homelessness, or substance abuse and medical conditions like trauma, chronic medical diseases, and an altered mental state [2–9]. Acute intoxication as the most likely reason is cited in 30%–79% of hypothermic deaths [2,9–11]. In many cases a combination of these conditions leads to the helpless situation. As stated, age of the very young or very old is one of the most important risk factors. While in infants the body mass is small and the surface is relatively large, loss of heat is faster than in adults. In small infants there is a complete dependence on others, especially regarding clothing and orientation [3]. In the elderly, lack of mobility, chronic cardiac and respiratory diseases, and dementia are common. Individuals with dementia are especially in danger because of their decreased capacity of orientation and estimation of dangerous conditions [2,4].
Morphological signs of death from the cold are given in literature. [2,3,7,8,11–16] (Table D3.9.2.1). One of the remarkable peculiarities in the death scene in cases of hypothermia is the irrational behavior in agony: like paradoxical undressing and uncoordinated movements. In these cases commonly sexual offense due to the undressing or influence of other persons due to trails according to falling and creeping of the hypothermic person is suspected (Figure D3.9.2.1, Table D3.9.2.2). These sign of irrational behavior could be found in 10%–82 % of cases of death from hypothermic [2,17]. All these findings are merely a sign of exposure to the cold but not of death from it, so the declaration of the manner of death remains difficult: Natural, accidental, suicidal, and even homicidal should be taken into account [6,18]. While the classification of accidental is common in death from hypothermia, suicidal and homicidal are rare and difficult to verify. Huusko and Hirvonen [8] mentioned the problem that in absence of a farewell letter, it is nearly impossible to verify if an intoxication in a cold environment
FIGURE D3.9.2.1 Paradoxical undressing in a case of fatal hypothermia.
© 2009 by Taylor & Francis Group, LLC
356
The Virtopsy Approach
FIGURE D3.9.2.2 Wischnewski’s spot-like bleedings to the gastric mucosa with small hematinized lesion with a maximum diameter of 0.2 cm.
was accidental or deliberately induced with the intention of freezing to death. The highest diagnostic value in medicolegal investigation seems to be Wischnewski’s spot-like bleedings of the gastric mucosa, which are described in 14% to 100% of cases of death from hypothermia, followed by purple-red blotches of the skin in nonhypostatic prominent areas (19%–90%) (Figure D3.9.2.2) [2,3,7,10,11,13,15,16]. The heterogenous group of pancreatic lesions includes focal or diffuse pancreatitis with or without hemorrhages as well as fat necrosis and vacuoles and is given in hypothermic death in percentages from 3% to 67% in literature [13,14], while muscular bleeding are still mentioned only infrequently in the medicolegal literature [17,18]. In radiological literature, studies on fatal hypothermia are rare and often related to traumatic injuries in outdoor extreme temperature conditions [19]. Postmortem radiological studies in the diagnosis of fatal hypothermia are still in progress, and to date only some of the internal findings are securely detectable. While presently the spot-like bleedings in the gastric mucosa remain too small to be visualizable by magnetic resonance imaging (MRI) or multislice computed tomography (MSCT) and pulmonary edema is too unspecific for the diagnosis of hypothermia, the hemorrhages to the muscles of the body core can be diagnosed by MRI (Figure D3.9.2.3a and Figure D3.9.2.3b). These findings can easily be located by MRI even in body structures that are not routinely dissected during autopsy. The problem in the radiological detection of small spotlike gastric bleedings—typical in hypothermic death—by MRI might be the limitation due to the technical development in high-field Tesla scanners. In ex vivo studies with high Tesla MRI, a differentiation of the layers of the gastric wall was possible [20,21]. The visualization of spot-like
© 2009 by Taylor & Francis Group, LLC
FIGURE D3.9.2.3 Findings in a 53-year-old man who was found dead outdoors in winter. Traumatic origin of muscle bleedings were not found in radiological examination or in autopsy clues. Especially the overlying skin and fatty tissue did not show any signs of mechanical impact in this area: (top) Axial cross-section (short tau inversion recovery [STIR], General Electric echo speed 1.5 T) of the pelvis. In the left gluteal muscle and in the inner pelvic muscles on both sides hemorrhages could be detected (red arrows). (bottom) Autopsy finding of the left gluteal muscle with an intramuscular hemorrhage.
superficial lesion should be possible in the future when highfield scanners for the whole body might be available. D3.9.2.1 References 1. Centers for Disease Control and Prevention. 2005. Hypothermia-related death—United States, 2003–2004. MMWR 54:173–75. 2. Oesterhelweg, L., H. Klotzbach, and K. Püschel. 2004. Epidemiological and phenomenological aspects of death from hypothermia. In Hypothermia. Clinical, Pathomorphological and Forensic Features, ed. M. Oehmichen, 105–13. Lübeck: Schmidt-Römhild. 3. Saukko, P. and B. Knight. 2004. Knight’s Forensic Pathology. London: Arnold.
Incident-Specific Cases
4. Rango, N. 1985. The social epidemiology of accidental hypothermia among the aged. Gerontologist 25:424–30. 5. Boswick, J. A., J. D. Thompson, and R. A. Jonas. 1979. The epidemiology of cold injuries. Surg Gynecol Obstet 149:326–32. 6. Huusko, R. and J. Hirvonen. 1988. The problem of determining the manner of death as suicide or accident in borderline cases. Z Rechtsmed 100:207–13. 7. Madea, B., J. Preuß, V. Henn, and E. Lignitz. 2004. Morphological findings in fatal hypothermia and their pathogenesis. In Hypothermia. Clinical, Pathomorphological and Forensic Features, ed. M. Oehmichen, 181–204. Lübeck: Schmidt-Römhild. 8. Schneider, V. and E. Klug. 1980. Tod durch Unterkühlung. Z Rechtsmed 86:59–69. 9. Woodhouse, P., W. R. Keatinge, and S. R. K. Coleshaw. 1989. Factors associated with hypothermia in patients admitted to a group of inner city hospitals. Lancet 2(8673):1201–05. 10. Krjukoff, A. 1914. Beitrag zur Frage des Todes durch Erfrieren. Vjschr Gerichtl Med 47:79–101. 11. Birchmeyer, M.S. and E. K. Mitchell. 1989. Wischnewski revisited. The diagnostic value of gastric mucosal ulcers in hypothermic death. Am J Forensic Med Pathol 10:28–30. 12. Dirnhofer, R. and T. Sigrist. 1979. Muscle hemorrhages in the body core: asign of a vital reaction in death from hypothermia? Beitr Gerichtl Med 37:159–66. 13. Hirvonen, J. 1976. Necropsy findings in fatal hypothermia cases. Forensic Sci 8:155–64. 14. Preuss, J., E. Lignitz, R. Dettmeyer, and B. Madea. 2007. Pancreatic changes in cases of death due to hypothermia. Forensic Sci Int 166:194–98. 15. Sigrist, T., C. Markwalder, and R. Dirnhofer. 1990. Veränderungen der Skelettmuskulatur beim Tod durch Unterkühlung. Z Rechtsmed 103:463–72. 16. Wischnewsky, S. 1895. Ein neues Kennenzeichen des Todes durch Erfrieren. Bote Gerichtl Med 3:12. 17. Rothschild, M. A. and V. Schneider. 1995. “Terminal Burrowing Behaviour”—a phenomenon of lethal hypothermia. Int J Leg Med 107:250–56. 18. Cala, A. D. and C. H. Lawrence. 2001. Suspicious circumstances and inexplicable wounds do not a murder make. MJA 175:621–22. 19 Grosse, A. B., C. A. Grosse, L. S. Steinbach, H. Zimmermann, and S. Anderson. 2007. Imaging findings of avalanche victims. Skeletal Radiol 46:515–21. 20. Sato, C., S. Naganawa, H. Kumada, S. Miura, and T. Ishigaki. 2004. MR imaging of gastric cancer in vitro: accuracy of invasion depth diagnosis. Eur Radiol 14:1543–49. 21. Yamada, I., N. Saito, K. Takeshita, et al. 2001. Early gastric carcinoma: evaluation with high-spatial-resolution MR imaging in vitro. Radiology 220:115–21.
D3.10 ELECTRICITY Danny Spendlove, Kimberlee Potter, and Michael J. Thali
357
TABLE D3.10.1 Effects of Current on the Body Effect of frequency Thermal effects Sensation Motor function Cardiac function Nerve function Burns
ion concentration difference rather than by electromagnetic induction as in metallic conductors. Electrical characteristics can be measured for diagnostic purposes as in cardiology or neurology, or the body can be a part of an electrical circuit in cases of surgery (electrocoagulation), where the current flow is controlled to prevent the body from injury [1–5]. The human body can also be subject to uncontrolled electrical currents or lightning shocks. The effects (Table D3.10.1) on a human body can vary from a sparking feeling to burns and death. The severity of tissue damage is directly related to a number of physical factors, which include current, voltage, resistance, and time [2–5]. For biological damage to occur, the body must be incorporated into an electrical circuit to facilitate passage of electrons through the different tissues. The patterns of current flow are a result of characteristics of the different tissues and the geometry of the structures in the path of the actual current. It tends to take the shortest route between entry and exit [2,3,5]. This path is important in determining its effects as different tissues and structures have different susceptibilities to the effects. It has been claimed that the most dangerous is contact with the right hand and exit through the feet, as this causes the current to pass obliquely along the axis of the heart. Intact skin has a complex voltage-dependent impedance that is affected by several factors such as skin thickness, temperature, and moisture of the skin. As the body contact of the current flow increases heating, tissue damage can occur as a result of resistance fall to a low value in order to allow higher currents to flow unimpeded. According to Odell [5], the effects of the current on the body can be classified to the following actions: r r r r r r r
Effect of frequency Thermal effects Sensation Motor function Cardiac function Nerve function Burns
D3.10.1 INTRODUCTION
D3.10.1.1 Effect of Frequency
Physiological processes are often associated with electrical activity, although they are often merely artifacts of biological processes. This electrical activity is normally generated by
Alternating current (AC) is more dangerous than direct current (DC) and has a more pronounced effect on muscle, nerve, and sensation function [3–5].
© 2009 by Taylor & Francis Group, LLC
358
The Virtopsy Approach
D3.10.1.2 Thermal Effects
D3.10.1.6 Nerve Function
Because the high resistance of the skin causes an energy transfer from the electron flow to the skin, electrical burns on the skin occur. High current densities in small volumes will give greater heating effects and a thermal concentration responsible for greater tissue damage. The temperature in the tissues directly under the contact point can easily reach 95nC. Tissue damage can occur within 25 seconds when the temperature reaches a mere 50nC [1,4,5].
Nerve tissue is also an electrical conductor, which responds in similar fashion to electrical stimulus, leading to muscle spasm or sensory effects like blocking the stimulus causing a reversible paralysis that can last for hours. Shocks to the nervous system can cause convulsions or unconsciousness, with cardiac effects when there is minimal involvement of the heart in the current path [5–8]. D3.10.1.7 Burns
D3.10.1.3 Sensation Pleasant or unpleasant exaggerated sensations mimic can be caused by electrical stimulation of the sensory nerves. The sensory threshold and intensity depend on the contact site and are related to current changes and frequency. As the frequency becomes higher, the sensory system becomes less able to respond to the rapid changes, and thus the pain sensation becomes less [4,5–8].
Electrical burns are typically indistinguishable from thermal burns of other causes. The current heats up the tissue fluids and produces steam, which can split the layers of the skin and produce a raised blister. This blister can rupture if the current continues or cool down and collapse. The burn lesion can have an impression mark of the applying conductor. The collapsed blister is often annular with a white, gray, and pale area with an umbilicated center. In high voltage, the mark takes the appearance of a hard brownish nodule, usually raised above the surrounding surface [1–5,18].
D3.10.1.4 Motor Function When electricity is applied to muscle tissue, it will result in contractions or twitching or in a sustained contraction (i.e., spasm or tetanic contraction), depending on the pulse frequency. The commercial AC power is high enough to produce tetanic spasm. These spasms can lead to bone fractures, paralysis of the respiratory muscles, or arrhythmias of the heart, causing a ventricular fibrillation [1–12].
D3.10.1.5 Cardiac Function As an “electrically” dependent organ, the heart is affected by externally applied currents, depending on the magnitude, duration, and timing of the current with respect to the cardiac cycle. This could result in disturbances in cardiac electric conduction, leading to extrasystole or other arrhythmias like ventricular fibrillation. Acute death usually results from ventricular fibrillation induced by disruption of coordinated conduction through the cardiac muscle [13,14]. Another cause of death is the respiratory arrest due to tetanic spasm [5–8]. Late cardiovascular effects of an electrical shock include acute myocardial necrosis, myocardial ischemia (without necrosis), heart failure, arrhythmias, hemorrhagic pericarditis, acute hypertension (hypertensive crisis) with peripheral vasospam, and anomalous electrocardiogram (ECG) changes. These late complications can occur several months after the incident. The first indications for cardiac complications are high-level serum myoglobin and other heart enzymes [12–17]. In cardioversion or defibrillation, the heart is subjected to a single high-voltage shock sufficient to simultaneously depolarize most or all of the myocardium, which may lead to asystole or an opportunity for spontaneous cardiac activity.
© 2009 by Taylor & Francis Group, LLC
D3.10.2 RISK FACTORS Factors that have some influence on the current flow, or influence electrocution, are wet surroundings or body and a high air humidity, as seen in accidents where defective machines (e.g., washing machines with a leakage) are the origin for electrocution deaths [11,18,19]. Decreasing the skin resistance as a result of wet skin or prolonged current contact might increase current intensities and cause injuries to the heart and the inner organs. In a wet surrounding, burning marks may be absent [7]. Of course, preexisting cardiac anomalies potentially increase the risk of death due to an electric shock. With preexisting cardiac arrhythmias, the normal rhythm is easily changed into a (deathly) arrhythmia when a current flow goes through the chest [20–22]. The current flow is a hotly discussed item. Some authors have noticed that more victims died due to a current flow from arms to the legs, although others describe a more lethal flow from the upper to lower extremities [12,13,23].
D3.10.3 FORENSIC FINDINGS Typical morphological findings can be sparse or absent in electrical injuries, although excitable tissue can have specific findings due to the path of the current flow (See Table D3.10.2). The occurrence of typical findings makes the diagnosis of electrocution easy, but the absence can lead to considerable problems. For example, homicides may remain undiscovered, or defective machines can lead to further deaths [2,3]. Electrical burns or current marks will appear mostly in high-voltage electricity injuries but can be absent in cases of low-voltage accidents [1–4,24,25]. They can even be absent
Incident-Specific Cases
359
TABLE D3.10.2 Typical Morphological Findings Sparse or absent Tissue can have specific findings due to the path of the current flow Electrical burns, current marks Typical burn lesions in surrounding tissue (see Chapter D3.9.1, “Thermal Damage: Heat and Burns”) Earthing, grounding lesions (Internal) current mark Sometimes petechial hemorrhages Heart (histological) disorders Electrocution fractures
in cases where a big area of entrance and a minimal resistance is involved, like suicides with a hair dryer in a bathtub. The electricity flowing through the body can cause the onset of rigor mortis to appear more quickly due to contraction of the muscle and depletion of adenosine-tri-phosphate (ATP) [1–3,26]. Typical places for current marks are the flexor sides of the fingers and feet and the tips of the fingers. They may appear either as an erythematous area of blistering or as irregular chalky white lesions with or without raised borders—due to the forming of blisters caused by the generated steam—and a central crater, sometimes penetrating quite deeply into the skin. The blanched skin could be caused by vasospasm of the vessel musculature and could be interpreted as virtually pathognomonic of electrical injury [1–4]. Where contact is less firm, such as with an air gap between conductor and skin, the current is able to jump as a spark. This can cause melting of skin keratin into a hard brownish nodule when cooled. With high voltage (multikilovolt range), sparking may occur over a larger distance, causing multiple spark lesions and giving the skin a “crocodile” effect [4,27]. In the direct surroundings of the current marks, typical burn lesions can be found, especially when high voltage is involved, when the conductor is large, or when the contact time is prolonged. As described before, the bleached skin area is caused by the creation of steam and the formation of blisters, as seen in thermal burns caused by cooking. With prolonged exposure to electricity, the burns become more severe, causing third-degree burning wounds in the direct surroundings to the electrical mark [1–5,25,27,28]. When 240 V, like domestic voltage, are applied over a prolonged time and the path of the current is through the organs, like in suicides where the conductors are applied to the sides of the abdomen, organs and tissues can be cooked and deep muscular damage is found. The skin can have extensive peeling or blistering. In a short current application time, internal findings may be spare or absent because of the aqueous organs and conductive electrolytes [24–28]. A postmortem search for the exit marks, called earthing or grounding lesions, has to be performed, although this may be difficult (Figure D3.10.1a). Places of predilection are the
© 2009 by Taylor & Francis Group, LLC
FIGURE D3.10.1A A 35-year-old woman was found dead next to a ladder under a cherry tree. Current mark on digit 5 of the left foot.
contralateral hand or the feet. Often no exit wounds are found. Sometimes an internal current mark is found at autopsy, with typical histological findings, like hypercontraction bands of the muscular tissues and coagulative changes of the nervous system, although it is not clear yet whether these internal marks are formed antemortem or postmortem [1–5]. The current mark itself has typical findings, as described by Schäffner [29] in 1965, with subepidermal or intraepidermal bullae, honeycomb-like changes in cells, and palisade-like or stream-like clustering of elongated nuclei of the stratum basale. Krager et al. [30] described petechial hemorrhages in 2002. In 74% of the cases petechial hemorrhage was seen, with predilection places of the eyelid skin and conjunctivae, the visceral pleura, and epicard. In cases where no petechial hemorrhages were found, a large contact surface was present, although it is not standard in cases of high voltage. The hemorrhages can be seen in asphyxial types of electrocution or as caused by massive congestion due to cardiac arrest and can be seen as a vital sign. As described by several authors [12–17,31–35] numerous complications of the cardiovascular system are triggered on the onset of the electricity. The effects documented include myocardial ischemia with and without necrosis, arrhythmias, heart failure, acute hypertension with vascular peripheral spasm, hemorrhagic pericarditis, and nonspecific ECG changes. The damage to the conducting system of the heart may lead to complication several months after the incident, which can be clear from complaints in the history taking and the use of several diagnostic tools (e.g., ECG, echocardiography, scintigraphy) and blood examination concerning the cardiac isoenzyme levels (i.e., creatine phosphokinase [CPK], CPK-MB, myoglobine) [13,32–35]. Morphologically contraction band necrosis, myocytolysis, striation loss, and nuclear disappearance have been observed. Fineschi et al. described the myocardial fiber breakup (MFB), which include the following:
360
The Virtopsy Approach
1. Bundles of distended myocardial cells, alternating with hypercontracted ones, in which the myocardial nuclei have, instead of an ovoid form, a square morphology. There can be a segmentation of the intercalated discs. 2. Hyperdistended myocardial cells are alternated with hypercontracted cells. 3. The bands of hypercontracted sarcomeres with stretched, often separated ones [13,14]. Electrocution fractures can be seen in low-voltage as well as high-voltage electrical accidents due to the muscles contractions caused by the flowing current. These muscular-induced fractures of the bones were frequently seen before the use of relaxants in shock and electroconvulsive therapy [10,11].
D3.10.4 RADIOLOGICAL FINDINGS Computed tomography (CT) scans are used in several postmortem examinations to assess and evaluate skeletal damage, internal organ injuries, and pathology as well as pathologic gas collection and embolism [36–39]. Magnetic resonance imaging (MRI), despite its clinical utility in examining softtissue trauma, organ injuries, and gross pathology, has not been widely exploited by the pathology community [40–42]. Because the external findings in autopsy show no relevant changes to the body, except to the aforementioned current mark, whole-body CT scans show no changes in the body. In many other cases, however, the resolution of the clinical CT and MRI scans was not sufficient to answer questions regarding forensic wound analysis, although minor changes are visible in MR (Figure D3.10.1b and Figure D3.10.1c). Emerging technologies, such as high-resolution CT (micro-CT) and magnetic resonance microscopy (MRM or micro-MR), may provide the relevant information and the requisite resolution [43,44].
FIGURE D3.10.1C MRI T1 image of the left foot, axial view. The current mark is visible in the circle.
In a study by Thali et al. [36], electrical entrance and exit marks were studied with micro-MR and were documented against histologic sections (Figure D3.10.2 and Figure D3.10.3). For this MR examination, intact skin samples were rehydrated and imaged in phosphate-buffered saline. The experiments were performed on a spectrometer, coupled to a horizontalbore magnet operating at 7 T (300 MHz for protons). The skin samples were imaged in a 35 mm probe at room temperature
A
© 2009 by Taylor & Francis Group, LLC
L2
B
L3
C
FIGURE D3.10.1B MRI T2 image; the current mark is visible (circle).
L1
D
FIGURE D3.10.2 Electric exit wounds on the (A) left and (B) right feet are indicated with white arrows. Lesion L1 on the small toe of the left foot was classified as a third-degree burn because of extensive charring of surrounding tissues. L2 on the left foot, composed of two epidermal blisters, was classified as a second-degree burn because of partial thickness dermal necrosis observed at the center of the wound site. L3 on the right foot with a central epidermal defect and full thickness dermal necrosis was also classified as a second-degree burn. The excised skin specimen of L3 is shown in (C) with the edges of the epidermal blister retracted to show the dermis. Carbonized dermal tissue (dark) and heat-coagulated dermal collagen (light brown) can be observed in the unstained tissue section of L3 shown in (D).
Incident-Specific Cases
A
361
Central
Interm ediate
Peripheral A
B
C B
C D Peripheral Zone
FIGURE D3.10.3 (A) Hematoxylin and eosin (H&E)-stained histologic section corresponding to the unstained tissue section shown in Figure D3.10.2(d). The central, intermediate, and peripheral zones of the L3 wound site are identified. The red arrow indicates the carbonized area, and the black arrows indicate the occluded vessels in the central zone. (B) H&E section at higher magnification (magn. 40 x) showing elongated cellular structures at the epidermal–dermal junction at the periphery of the wound site (white arrows). (C) Histologic section stained with Movat pentachrome with carbonized area indicated with a red arrow.
with fat suppression, reducing the proton signals coming from the subcutaneous fat. Two-dimensional images acquired perpendicular to the skin surface and directly through the wound site were used to measure the MR properties for the various tissue zones. Each 2D image had a slice thickness of 2 mm and an in-plane resolution of 120 Mm. Three-dimensional images of intact skin specimens were acquired with a rapid acquisition with relaxation enhancement (RARE) imaging sequence (repetition time [TR]/echo time [TE] 2000/8 ms; number of excitations [NEX] 1; RARE 8) at 150 Mm isotropic resolution. Water proton transverse relaxation times (T2) and water proton density (PD) values at each pixel were calculated on a pixel-by-pixel basis from 16 images acquired with a multiecho sequence (TR/TE 5000/12 ms). The PD values, which were extracted by this calculation, provide a direct measure of the number of water protons present in each pixel in the image. Assuming the PD values for saline represent 100% hydration, then the PD values for the tissue when normalized to that of saline yield the hydration state of the tissue. By comparison, T2 values for water protons in the tissue yield an indirect measure of the mobility of water within the tissue. Thus, very mobile water molecules, such as those in saline, will have the highest T2 values [45,46]. Magnetization transfer (MT) maps were calculated on a pixel-by-pixel basis using the following equation: 1 Mso/Mo where Mso/Mo gives the ratio of pixel intensities from images acquired with and without the application of a 5-s, 12-MT
© 2009 by Taylor & Francis Group, LLC
Central Zone
FIGURE D3.10.4 (A) Sirius red-stained histological sections through the peripheral (left), intermediate (middle), and central (right) zones of the L3 wound site. PD (protein density) (B), T2 (C), and MT (magnetization transfer) (D) maps from the same approximate locations are also shown.
saturation pulse 6000 Hz off resonance. This parameter measures the rate of transfer of magnetization from mobile water protons to protons on large immobile macromolecules. For collagen-containing tissues, MT is very efficient (high MT value), and for saline it is nonexistent (MT is zero). To assess the severity of the tissue damage at the wound site, 2D PD, T2, and MT maps were acquired perpendicular to the skin surface and directly through each current mark (Figure D3.10.4). The reported PD, T2, and MT values represent the mean o standard deviation (SD) of pixel values for different regions of interest.
D3.10.5 MICRO-MR FINDINGS For normal skin in the peripheral zone, at least four layers of skin were resolved in the PD map. At the surface, there was the stratum corneum, which was visible in this experiment because it had become hydrated during the fixation process. Below the stratum corneum was the more hydrated (brighter) stratum spinosum. The next layer was the dermis, which was less hydrated (darker) and more heterogeneous because of the presence of cutaneous appendages such as sweat glands and hair follicles. Below the dermis was the hypodermis. The hypodermis was dark in the MR images shown because it was composed mostly of fat and the signal from fat was suppressed during image acquisition. In the intermediate zone, the dermis was much brighter, with higher PD (78 o 8%) and T2 (40 o 9 ms) values compared with normal dermal tissue (PD 76 o 3%, T2 30 o 7 ms). The bright area seen on both the PD and T2 map extended into the superficial hypodermis, comparable to the edematous area. In contrast, the dermis in the central zone was much darker on T2 and PD maps compared with normal dermal tissue. This area correlated well with the thermally
362
A
The Virtopsy Approach
B
site. This lesion might be attributed to Joule heating by proximal osseous tissue.
D3.10.6 CONCLUSION
C
FIGURE D3.10.5 (A) T2-weighted cross-sectional image of the L3 exit wound showing electric injury pattern with carbonized central zone (red arrow) and edematous intermediate zone with thrombosed vessel (white arrow). (B) T2-weighted image showing the course of the occluded vessel, the dark carbonized area, and the bright intermediate zone. Arrows show proposed current path. (C) 3D stereogram showing occluded blood vessels and area of edema segmented from the 3D MR data. The cloudy area of edema is green and the areas of occlusion are red. The arrows denote the proposed path of the current.
damaged tissue. Importantly, carbonized tissue had the lowest signal intensity and much reduced PD (30 o 15%) and T2 (14 o 9 ms) values. On MT maps, bright pixels occurred in regions of collagenous tissues such as the dermis, vessel walls, and connective tissue septa in the hypodermis. Compared with normal dermis in the peripheral zone (MT 0.70 o 0.03), thermally damaged dermis, predominantly in the central zone, had a higher MT value (0.74 o 0.07), and dermal edema in the intermediate zone had a lower MT value (0.65 o 0.04). These results suggest that water proton MRM parameters can be used to assess the extent and severity of electric exit wounds. Accordingly, T2-weighted images (Figure D3.10.5(a,b)) extracted from a 3D data set revealed the extent of severe thermal damage seen as reduced signal intensity and the extent of edema seen as an enhancement of signal intensity. Edema extended into the dark hypodermal layer consistent with histologic findings. The hypodermal layer was dark except in regions of edema, in the lumen of blood vessels running parallel to the skin surface, and in connective tissue septa between clusters of adipocytes. There was a focal reduction in image intensity in the vessel lumen compared with other vessels present in the hypodermis. The occluded vessels and the area of edema were segmented interactively from the 3D MRM image and rendered as a stereogram (Figure D3.10.5(c)) to demonstrate the proposed path of the current along the vessel in three dimensions. Volumetric images of an intact lesion also revealed a focal lesion, characterized by the disorganization of the connective tissue septa, deep in the hypodermis but distal to the wound
© 2009 by Taylor & Francis Group, LLC
Although CT and MR scans are widely used in clinical settings and hardly used in the office of forensic pathologists, CT and MRI are useful technical equipment for examining corpses of many forensic cases. With a short current exposure time no marks of current flow have to visible, although an entrance or exit mark might be seen. With prolonged exposure time to the electricity, an internal current mark might be found. Conventional CT or MR examinations cannot yet reveal the current path, although, as seen in histological examinations of the current marks, micro-MR can make the typical lesions visible for further examination.
D3.10.7 REFERENCES 1. DiMaio, V.J. and D. DiMaio. 2001. Electrocution in Forensic Pathology, 2d ed. CRC Press, London. 2. Saukko, P. and B. Knight. 2004. Knight’s Forensic Pathology, 3d ed. Arnold Publishers, Oxford University Press, New York. 3. Brinkmann, B. and B. Madea. 2004. Handbuch gerichtliche Medizin [Handbook of forensic medicine, authors’ translation]. Springer Verlag, Berlin. 4. Reimann, W. and O. Prokop. 1985. Vademecum Gerichtsmedizin, 4th ed. [Vademecum forensic medicine, authors’ translation]. VEB Verlag Volk und Gesundheit, Berlin. 5. Odell, M. 1997. The human body as an electric circuit. J Clin Forensic Med 4, 1–6 6. Patten, D.M. 1992. Lightning and electrical injuries. Neurol. Clin. 10(4): 1047−58. Review. 7. Siroofsky, M.D., R.Y. Hawley, and H. Manz. 1991 Progressive motor neuron disease associated with electrical injury. Muscle Nerve 14(10): 977−80. 8. Alexander, L. 1941. Electrical injuries of the nervous system. J Nerv Ment Dis 94, 622–32. 9. Dumas, J.L. and N. Walker. 1992. Bilateral scapular fractures secondary to electrical shock. Arch Orthopaed Trauma Surg 111(5):287–88. 10. Tarquinio, T., M.E. Weinstein, and R.W. Virgilio. 1979. Bilateral scapular fractures from accidental electric shock. J Trauma 19(2):132–33. 11. Bailey, D., S. Forget, and P. Gaudreault. 2001. Prevalence of potential risk factors in victims of electrocution. Forensic Sci Int 123:58–62. 12. Kouwenhoven, W.B., D.R. Hooker, and O.R. Langworthy. 1932. The current flowing through the heart under conditions of electric shock. Am J Physiol 100:344–50. 13. Fineschi, V., S. Di Donato, S. Mondillo, and E. Turillazzi. 2006. Electric shock: cardiac effects relative to nonfatal injuries and post-mortem findings in fatal cases. Int J Cardiology 111:6–11. 14. Fineschi, V., S.B. Karch, S. D’Errico, C. Pomara, I. Riezzo, and E. Turillazzi 2006. Cardiac pathology in death from electrocution. Int J Legal Med 120:79–82. 15. Baroldi, G., M.D. Silver, M. Parolini, C. Pomara, E. Turillazzi, and V. Fineschi. 2005. Myofiber break-up: a marker of ventricular fibrillation in sudden cardiac death. Int J Cardiol 100:435–41.
Incident-Specific Cases
16. Fieguth, A., G. Schumann, H.D. Tröger, and W.J. Kleeman. 1999. The effect of lethal electrical shock on post-mortem serum myoglobin concentrations. Forensic Sci Int 105:75–82. 17. Zack, F., U. Hammer, I. Klett, and R. Wegener. 1997. Myocardial injury due to lightning. Int J Legal Med 110:326–28. 18. Gunwantrao Wankhede, A. and D.R. Sariya. 2006. An electrocution by metal kite line. Forensic Sci Int 132:141–43. 19. Di Nunno, N., L. Vimercati, L. Viola, and F. Vimercati. 2003. A case of electrocution during illegal fishing activities. Am J Forensic Med Pathol 24:164–67. 20. Herlevsen, P. and P.T. Andreasen. 1987. Constitutional predisposition to vasovagal syncope: additional risk factor in patients exposed to electrical injuries? Br Heart J 57(3):284–85. 21. Jackson, S.H. and D.J. Parry. 1980. Lightning and the heart. Br Heart J 43(4):454–57. 22. Lichtenberg, R., D. Dries, K. Ward, W. Marshall, and P. Scanlon. 1993. Cardiovascular effects of lightning strikes. J Am Coll Cardiol 21(2):531–36. 23. Anders, S., J. Matschke, and M. Tsokos. 2001. Internal current mark in a case of suicide by electrocution. Am J Forensic Med Pathol 22(4):370–73. 24. Nichter, L.S., C.A. Bryant, J.G. Kenney, et al. 1984. Injuries due to commercial electric current. J Burn Care Rehabil 5:124–37. 25. Zhang, P. and S. Cai. 1995. Study on electrocution death by low-voltage. Forensic Sci Int 76(2):115–19. 26. Hunt, J.L., A.D. Mason, T.S. Masterson, and B.A. Pruitt. 1976. The pathophysiology of acute electric burns. J Trauma 16:335–40. 27. Fish, R. 1993. Electric shock: part II. Nature and mechanisms of injury. J Emerg Med 11(4):457–62. 28. Thomsen, H.K., L. Danielsen, O. Nielsen, et al. 1981. Early epidermal changes in heat- and electrically injured pig skin. I. A light microscopic study. Forensic Sci Int 17:133–43. 29. Schäffner, M. 1965. Untersuchungen über Histologie und Metallisation nach elektrischen Einwirkungen auf die Haut. Dtsch Z Ger Med 56:269–80. 30. Karger, B., O. Süggeler, and B. Brinkmann. 2002. Electrocution—autopsy study with emphasis on “electrical petechiae.” Forensic Sci Int (126):210–13. 31. Kleiner, J.P. and J.H. Wilkin. 1978. Cardiac effects of lightning stroke. J Am Med Assoc 240(25):2757–59. 32. Arya, K.R., G.K. Taori, and S.S. Khanna. 1996. Electrocardiographic manifestations following electric injury. Int J Cardiol 57(1):100–01. 33. Chandra, S.S., C.O. Siu, and A.M. Munster. 1990. Clinical predictors of myocardial damage after high voltage electrical injury. Crit Care Med 18(3):293–97. 34. Guinard, J.P., R. Chiolero, E. Buchser, et al. 1987. Myocardial injury after electrical burns: short and long term study. Scand J Plast Reconstructive Surg Hand Surg 21(3):301–02. 35. Xenopoulos, N., A. Movahed, P. Hudson, and W.C. Reeves. 1991. Myocardial injury in electrocution. Am Heart J 122(5):1481–84. 36. Thali, M.J., R. Dirnhofer, R. Becker, W. Oliver, and K. Potter. 2004. Is “virtual histology” the next step after the “virtual autopsy”? Magnetic resonance microscopy in forensic medicine. Magn Reson Imaging 22(8):1131–38. 37. Thali, M.J., W. Schweitzer, K. Yen, et al. 2003. New horizons in forensic radiology: the 60-second “digital autopsy”; fullbody examination of a gunshot victim by multi-slice computed tomography. Am J Forensic Med Pathol 24:22–27.
© 2009 by Taylor & Francis Group, LLC
363
38. Jackowski, C., M.J. Thali, M. Sonnenschein, et al. 2004. Visualization and quantification of air embolism structure by processing postmortem MSCT data. J Forensic Sci 49(6):1339–42. 39. Aghayev, E., K. Yen, M. Sonnenschein, et al. 2005. Pneumomediastinum and soft tissue emphysema of the neck in post-mortem CT and MRI; a new vital sign in hanging? Forensic Sci Int 153:181–83. 40. Bisset, R.A., N.B. Thomas, I.W. Turnball, and S. Lee. 2002. Postmortem examination using magnetic resonance imaging: four year review of a working service. Br Med J 324:1423–24. 41. Patriquin, L., A. Kassarjian, M. Barish, et al. 2001. Postmortem whole-body magnetic resonance as an adjunct to autopsy: preliminary clinical experience. J Magn Reson Imaging 13:277–87. 42. Thali, M.J., K. Yen, W. Schweitzer, et al. 2003. Virtopsy, a new imaging horizon in forensic pathology: virtual autopsy by postmortem multislice computed tomography (MSCT) and magnetic resonance imaging (MRI)—A feasibility study. J Forensic Sci 48:386–403. 43. Hennig, J., A. Nauerth, and H. Friedburg. 1986. RARE imaging: a fast imaging method for clinical MR. Magn Reson Med 3:823–33. 44. Thali, M.J., U. Taubenreuther, M. Karolczak, et al. 2003. Forensic microradiology: Micro-computed tomography (micro-CT) and analysis of patterned injuries inside of bone. J Forensic Sci 48(6):1336–42. 45. Ceckler, T.L., S.D. Wolff, V. Yip, S.A. Simon, and R.S. Balaban. 1992. Dynamic and chemical factors affecting water proton relaxation by macromolecules. J Magn Reson 98:637–45. 46. Kim, D.K., T.L. Ceckler, V.C. Hascall, A. Calabro, and R.S. Balaban. 1993. Analysis of water-macromolecule proton magnetization transfer in articular cartilage. Magn Reson Med 29:211–15.
D3.11 CLINICAL FORENSIC IMAGING K. Yen, R. Dirnhofer, and G. Ranner
D3.11.1 INTRODUCTION Clinical forensic medicine covers, among other issue, domestic violence, child abuse, sexual abuse, medical maltreatment, and traffic-related incidents [1]. Although the examination of living persons following accidental and nonaccidental trauma has a long tradition in legal medicine, a general political and social trend toward notification and processing of clinical forensic issues is currently observed [2]. Obviously, due to the growing sensitivity of the public, the number of clinical forensic expertises has increased significantly in many forensic units. This trend is reflected by the implementation of clinical forensic units with specialized personnel for examinations of living persons such as the establishment of a “forensic nursing” system in the United States in the past years. The 2007 meeting of the German Society of Legal Medicine also emphasizes the development of clinical forensic health-care centers. For obvious reasons, the application of invasive examination methods is limited in living persons; hence, the majority of internal body findings currently remain undetected at
364
forensic assessment. Forensic expertise is therefore based predominantly on the external findings in most cases. This underlines the need for an additional clinical forensic application of examination techniques such as magnetic resonance imaging (MRI) and computed tomography (CT), which have a certain potential for improving forensic investigation by offering noninvasive access.
D3.11.2 STANDARD FORENSIC EXAMINATION OF LIVING PERSONS Today’s forensic gold standard in the investigation of living persons is external examination of the body [3], which serves as the main basis for forensic expertise besides general case history and trace and toxicological analyses. By inspecting the whole body surface, all visible alterations and injuries are registered, documented, and finally interpreted. The external clinical forensic examination is sometimes, but not frequently, completed by further clinical investigation methods like laryngoscopy or laboratory analyses. Only if there is suspicion of injuries requiring medical treatment are radiological techniques or invasive examination methods applied [3]. Reflecting this, it becomes obvious that the clinical forensic gold standard is lacking in one major way: The internal findings generally escape today’s medicolegal assessment.
D3.11.3 RADIOLOGY IN CLINICAL FORENSIC MEDICINE Radiology plays an important role in the clinical evaluation of accidental and nonaccidental trauma, and extensive clinical experience exists with the application of imaging techniques for the assessment of traumatic injuries. Although the potential of radiology is indisputable in the clinical assessment of internal body findings and current research activities [4,5] indicate a great potential of the noninvasive radiological methods also in clinical forensic medicine, radiological studies have rarely been applied in living persons for forensic purposes only. A reason for this might be the limited experience of many forensic trained persons with the application and interpretation of clinical radiological imaging, especially regarding CT and MRI. The forensic assessor often has no direct access to clinical radiology units and therefore has to rely only on the records and remarks from the clinical history of the injured person. The limited use of radiological methods for medicolegal evaluation might also be due to some juridical and ethical issues that restrict the application of methods causing radiation exposure to living persons when these are applied only for forensic but not for clinical purposes. In the past years, conventional radiography was the method used predominantly for clinical forensic needs [6]. This was probably due to the fact that the availability of conventional radiography was generally superior to the other imaging methods, which are at present also associated with higher costs. Furthermore, the forensic experts were often accustane to conventional roentgen technology and only to a lesser degree to MRI and CT. Besides the routine application of x-ray techniques in hospitalized trauma cases, with
© 2009 by Taylor & Francis Group, LLC
The Virtopsy Approach
the clinical imaging data also serving as an additional basis for forensic expertise, the main clinical forensic issues for which conventional radiography was used was the analysis of foreign bodies [6] and, increasingly in recent years, age estimation [7]. Being easily accessible and affordable, x-ray is still frequently applied yet is being replaced more and more by CT in today’s clinical and also forensic practice. The development of the multidetector-array technique known as multislice CT (MSCT) allows CT to be performed much faster and more thoroughly than ever before. Excellent 2D and 3D reconstructions are available within minutes. This part of every standard examination substantially supports the visualization of relevant findings and issues that are especially important for forensic studies and documentation. The ability of CT imaging to depict traumatic injury in the clinical examination of living persons points toward the potential of the method regarding also the forensic assessment of injury. Especially the evaluation of bone lesions, traumatic organ changes, foreign bodies, and the presence and distribution of gas can be improved when it is based on CT examination. Being applied for specific clinical forensic purposes, the method has been used predominantly for the evaluation of gunshot injury [8,9] and age estimation [7]. MRI has demonstrated its obvious clinical forensic potential in a recent study by Yen et al. [5]. However, MRI has not found entrance into clinical forensic routine yet even though the method has some imaging properties that are of specific interest. As MRI offers excellent soft-tissue and organ contrast, the diagnosis of lesions that are hardly accessible using CT or x-ray as, for example, soft-tissue injuries caused by blunt impact or posttraumatic bone bruising, will become possible (Figure D3.11.1) [10]. Due to inflammation and healing processes in the living, such lesions are detected even superior to most postmortem forensic radiological examinations, especially if some hours lie between the incident and MRI. In the conventional clinical forensic assessment, this phenomenon is commonly known as flourishing of externally visible bruises, which tend to be visible better when external examination of the body is iterated few days after the primary postincidental evaluation. Specific MRI applications such as diffusion-weighted imaging (DWI) and diffusion tensor imaging (DTI), perfusion imaging, or functional MRI (fMRI) have a further potential to provide diagnostic improvements in clinical legal medicine. Despite its undoubted diagnostic value, MRI is rarely performed in the acute clinical management of trauma patients, mainly for medical and organizational reasons. The clinical forensic and postmortem MRI results from the Virtopsy Project [4,5,10–14] let us, however, suppose that MRI provides important forensic information that cannot be obtained using other imaging modalities; therefore, if the clinical status of the patient allows, the forensic application of MRI should be reflected when indicated (Table D3.11.1). The fact that MRI can be applied even days after the traumatic incident and still gives reliable information [5] is a clear advantage for forensic imaging, as the scans usually take place in a clinical routine MRI unit with all its organizational limitations that might
Incident-Specific Cases
A
365
C
D
B
E
FIGURE D3.11.1 The possibilities of modern imaging technologies CT and MRI are generally far beyond those of “classical” roentgenography. Especially MRI, which is not associated with radiation exposure, can bring benefit for the forensic assessment of living persons. From a diagnostic view, MRI offers unique possibilities of displaying soft-tissue, muscle, and ligament injuries as well as bone bruising and occult bone fractures. The presented case shows a comparison between roentgenography and MRI in a 41-year-old male who was hit by a car when riding his motorbike. Clinically, there was suspicion of a traumatic lesion of the right ankle joint; the roentgenogram, however, displayed no traumatic findings of the joint and foot (A). In contrast, sagittal T2-weighted MRI reveals fracture of the dorsal tibia (B, arrows), with edema of the bone fragment that appears slightly hyperintense. The tibio-fibular ligament is ruptured (C, frame), and distinct posttraumatic bone bruising is present in the second metatarsal and the medial cuneiform bone (D, frame). Besides the osseous and ligament injury, the subcutaneous fat-tissue lesions give important clues for the trauma mechanism that caused the injury: Contusion of the back and lateral rim of the foot is visible in (D, red arrows) and (E, frame), thus confirming the primary impact in this region and being consistent with massive plantar flexion and rotation of the foot that resulted in the injuries. The white arrows in (C) and (D) point toward T2-hyperintense edema that is present in the ankle region secondary to the sustained trauma: (A) Roentgenography. (B) Sagittal T2-weighted turbo spin-echo (TSE) sequence. (C) Transversal/oblique T2-weighted fat-saturated sequence. (D) Axial T2-weighted TSE sequence. (E) Coronal T1-weighted TSE sequence.
in some cases prohibit the examination within the first few hours after an incident. Being a highly subjective method regarding image acquisition and documentation and relying mainly on the experience of the examining person, the forensic value of ultrasound remains subject to further research. As the lack of reproducibility of ultrasound results principally deteriorates its use for medicolegal issues, its forensic application appears problematic. There are, however, several studies indicating some clinical forensic potential for the inexpensive and easily accessible method. Ultrasound has become an important clinical investigation tool, especially when it comes to evaluating fetal changes or blood flow and organ analyses, all of
© 2009 by Taylor & Francis Group, LLC
which could also be useful for forensic purposes. The forensic potential of sonography has already been shown for the assessment of skin and bone lesions [15–18], organ lesions [19], cardiac trauma [20,21], blood-flow measurements following strangulation [22–24], and body packing [25].
D3.11.4 FUTURE OF CLINICAL FORENSIC RADIOLOGY: POTENTIAL APPLICATIONS Several postmortem studies have demonstrated specific benefits that can be obtained by the application of CT and MRI in forensic cases, many of them described elsewhere in this
366
The Virtopsy Approach
TABLE D3.11.1 Possible Future Clinical-Forensic Indications of CT and MRI Incident Form
Forensic-Radiological Evaluation of Typical Case Relevant Findings1 CT*
Accidental injury
Strangulation Inflicted trauma
Traffic related injury: pedestrian
Bone injury, impact location, organ lesions
Traffic related injury: driver/occupant
Bone injury, impact location, organ lesions
Traffic related injury:others Falls
Bone injury, impact location, organ lesions Bone injury, impact location, organ lesions
Sports related accidental injury
Bone injury
Blunt force trauma
Laryngeal bone injury Bone injury, organ lesions
Gunshot incidents
Stabbings Child abuse
Shaken baby Battered child: blunt force trauma
Abuse of aged persons Torture Sexual assault Self-inflicted injury
Blunt force trauma Adults and children Psychiatric causes Insurance fraud
Body packing Medical maltreatment Intoxication Personal fitness Civil law issues
Fitness to drive, fitness to act etc. Short- and longtime trauma sequelae
Bone injury, bone entrance and exit wound, bullet particles, wound channel, gas embolism Wound channel, bone injury, gas detection, organ lesions Intracranial hemorrhage Recent or old skeletal injury ("babygram"), organ lesions Bone injury, organ lesions Sequelae of bone injury
Bone injury Foreign material Organ lesions, gas embolism,foreign body detection Cerebral status Organ status, previous diseases Skeletal and organ status
MRI** Soft tissue lesions, impact location, organ lesions, bone bruises Soft tissue lesions, seat belt injury,whiplash injury,organ lesions Soft tissue lesions, organ lesions Soft tissue lesions, impact location, organ lesions Soft tissue lesions, ligament structures, bone bruises Soft tissue lesions Soft tissue lesions, impact locations, organ lesions Wound channel lesions
Wound channel, organ lesions Soft tissue lesions, intracranial injuries Soft tissue lesions, organ lesions, bone bruises Soft tissue lesions, organ lesions (Extensive) scarring Accompanying soft tissue lesions Exclusion of blunt force injuries Soft tissue lesions Cerebral status: hypoxia, organ lesions Cerebral lesions, hypoxia issuesOrgan status, previous diseases Soft tissue and organ status, bone bruises
1 The apppearance offindings is strongly case-dependent. * CT is generally efficient for the evaluation of bone injuries, organ lesions and gas detection. ** MRI is predominantly suited for the diagnosis of soft tissue and organ lesions, as well as some special
book (see Chapter C4.3, Chapter C4.5, and Chapter D3). Even knowing that the radiological methods have a couple of specific restrictions that may cause diagnostic problems (e.g., limited image resolution or artifact), the clinical experience from years of MRI and CT imaging and the results of the preliminary forensic radiological imaging studies suggest a potential of the methods to improve mainly the following medicolegal issues: (1) the detection and classification of internal injuries (injury type and localization); (2) the evaluation of the extent and severity grading of internal injuries; (3) the improvement of forensic reconctruction (e.g., sequence of events, time-dependent behavior of injuries, determination of the danger to life or the fitness to act); and (4) the assessment of long-time sequelae of trauma. According to postmortem
© 2009 by Taylor & Francis Group, LLC
experience, especially CT and MRI of injuries following blunt force trauma with a subsequent reconstruction of the sequence of events appears to be advantageous for forensic expertise [4,10,11,14,26–29]. The postmortem acquired knowledge will serve as a profound basis for clinical forensic examinations where blunt force trauma is the most common injury type. A summary of potential future indications of MRI and CT is given in Table D3.11.1 and in the following sections. Future research will, however, be essential for ultimately defining the expedient indications of clinical forensic imaging as well as the application of the radiological techniques in forensic cases (e.g., specification of the applied methods and sequences for each indication) and the forensic interpretation of the imaging data.
Incident-Specific Cases
D3.11.4.1 Strangulation In cases of manual or ligature strangulation or in near-hanging incidents, clinical imaging of the neck is performed only when there is suspicion of neck injuries requiring medical treatment. In such cases CT is the method that is mostly applied [30,31]. Even if the forensic assessor has the possibility of using these clinical radiological CT data also for his or her expertise, the CT examination will usually not reveal the often discrete soft-tissue lesions that are associated with strangulation. In survived strangulation cases, the forensic examiner must always address the question as to whether an act has caused a life-threatening situation for the victim. In classical forensic examination this decision is mainly based on subjective data that are reported by the victim and other witnesses (e.g., loss of consciousness, loss of urine, hallucinatory phenomena, or other signs of cerebral hypoxia). The only objective sign that serves as a criterion indicating cerebral hypoxia is petechial hemorrhage of the conjunctivae, eyelids, oral mucosa, and facial skin [32]. Reflecting this, it becomes obvious that obtaining the inner strangulation findings that are well known from autopsy could have an important impact on the forensic and juridical treatment of these frequent cases. According to a recent clinical forensic study [5], MRI is well suited for the detection and documentation of the relevant neck findings in survived strangulation cases and also seems suited for extending the objective basis for the assessment of the danger to life (see below) (Figure D3.11.2, Figure D3.11.3, and Figure D3.11.4). For this reason, strangulation might become the first “routine” clinical forensic indication of MRI. MRI performed within hours to days after the incident (Figure D3.11.5) offers reliable results about the inner neck status following strangulation injury and therefore far extends today’s diagnostic basis. Within the virtopsy context at the University of Bern, clinical forensic MRI examinations have been performed in survived strangulation cases since 2001. The corresponding study [5] represents the first systematic approach toward the use of MRI for only forensic purposes, namely, to detect the victim’s internal neck injuries and to evaluate them in view of the interpretation of the danger to life. The study had a massive impact on the treatment of these cases at the forensic institute in Bern, with the local justice departments requesting MRI in all survived manual and ligature strangulation cases since the additional value of the radiological examination became obvious. Meanwhile, MRI of the neck has been applied in more than 40 survived strangulation cases in Bern. Based on the results of these examinations and the knowledge of the postmortem appearance of findings following strangulation, a preliminary approach has been established for appraising the danger to life. The carotid arteries and the upper airways were defined as the main “critical neck structures” where forcible compression for at least a few seconds is expected to result in cerebral hypoxia. Hemorrhage seen at MRI in close proximity to the critical neck structures was believed to prove forcible compression and therefore to indicate a high probability of cerebral hypoxia and a life-threatening situation for
© 2009 by Taylor & Francis Group, LLC
367
A
B
FIGURE D3.11.2 Subcutaneous hemorrhage on the right side of the neck (A, frame) and lymph node hemorrhage (A, arrows) in a woman following manual strangulation. The lymph nodes appear hyperintense at axial T2-weighted MRI. The victim reported loss of consciousness and hallucinatory phenomena. At the external examination, petechial hemorrhage was present in the conjunctivae. Using the imaging findings and the conventional examination results, the case had been stated as having caused danger to life. The left sternocleidomastoid muscle showed hyperintensity that was interpreted as hemorrhage at the image reading (B, T2-weighted axial MRI).
the victim during the act [5] (Figure D3.11.2). Hemorrhage of the neck lymph nodes was seen in excellent concordance with severe strangulation (verified by externally visible petechial hemorrhages), and the radiologically visible alterations (Figure D3.11.3; Figure D3.11.4) generally corresponded well to findings that were observed in a similar appearance in previous postmortem examinations [4] (see also Chapter D3.7, “Strangulation”). As forensic expertise demands in general, the whole set of case circumstances were taken into account for these analyses, and the expertise was not based on the MRI findings only [5]. Reflecting that the conventional forensic assessment of the danger to life in survived strangulation is based on only one
368
The Virtopsy Approach
A
A
B B
FIGURE D3.11.3 18-year-old male having suffered a forearm choke hold 38 hours before MRI. The axial gradient-echo MR image of the neck reveals distinct hyperintensity of the left sternocleidmastoid muscle (A, frame). The finding represents hemorrhage of the sternocleidoid muscle, which is situated near the “critical neck structures.” In accordance with the externally visible findings of extensive petechial hemorrhage in the eyelids and facial skin, the case was judged as having been a life-threatening event. At external examination, the left side of the neck presented with band-like and diffuse reddish discolorations (B).
objective finding (i.e., petechial hemorrhages), it becomes clear that visualizing the internal strangulation-related neck injuries by radiological imaging will increase the accuracy of forensic expertise in this field. Even though the results from the previously described clinical forensic radiological strangulation study yielded promising results, it is important to face the requirement for further research to establish examination and interpretation standards for the clinical forensic use of MRI and CT in strangulation incidents. D3.11.4.2 Accidental Trauma Several forms of accidental trauma are frequent subjects of forensic examination, most of them being traffic related or due to falls. Depending on the incident, the whole spectrum of traumatic injury can be found, including gunshot and heat-induced injuries, with the severity of the injuries varying from very discrete to severe and life-threatening lesions.
© 2009 by Taylor & Francis Group, LLC
FIGURE D3.11.4 In this woman, axial T2-weighted neck MRI revealed distinct hemorrhage in the soft tissues between the salivary gland and the base of the tongue on the right side (A, frame) and on the left side in close proximity to the larynx (B, frame). No petechial hemorrhages were present in this case at the external evaluation. The act was, however, interpreted as life-threatening, as the radiological neck findings supported the report of the victim that she had been forcedly strangled for some minutes, causing blackout.
Concerning traffic-related issues, for example, the investigation of persons with the focus on possible (medical) causes for an accident or the determination of “who was the driver” and other case-dependent reconstructive issues has to be provided by the forensic expert. Besides the clinical radiological examinations that are regularly available in cases who received hospital treatment, some issues might benefit from specific forensic imaging. Due to the fact that MRI and CT depict internal soft-tissue, organ, and bone lesions, the methods might improve the forensic reconstruction of traffic-related and other traumatic incidents by providing a sort of “internal impact pattern” (Figure D3.11.6, Figure D3.11.7, and Figure D3.11.8). In T2-weighted fat-suppressed sequences, MR, for instance, shows areas of high signal intensity in soft tissues where fluid or edema is accumulated. As in accidents the detection
Incident-Specific Cases
A
369
A
B
B
C
FIGURE D3.11.5 MRI of the neck has the potential to reveal caserelevant data even days after the incident. A 45-year-old woman reported localized pain on pressure on the left neck side 10 days after ligature strangulation. Axial MRI clearly depicted hemorrhage and swelling of the left sternocleidomastoid muscle (A, arrows). Lymph node hemorrhage (B, arrows, coronal slice) was also present. The radiological presumption that the victim had suffered severe strangulation could not be confirmed as no external findings were visible at the time the incident was reported.
of the impact site is generally of great importance, it might be useful to perform, for example, MRI and CT of the legs in pedestrians who were hit by a car to get information about the soft-tissue and bone lesions including bone marrow alterations or to apply the new technique of whole-body MRI (Figure D3.11.7) for obtaining a general injury survey. Another traffic-related and frequent finding is whiplash injury. Here, especially MRI might improve the assessment of objective findings that will help to underline or exclude the sequelae of whiplash. As described by Krakenes and Kaale [33–35], MRI is suited for revealing characteristic soft-tissue lesions such as alar ligament injury. Being a highly controversial issue [36], ongoing clinical studies will help also with the specific clinical forensic evaluation of the value of the imaging methods in this field. Regarding the differentiation of various forms of force injury, or the forensic highly relevant differentiation of accidental from inflicted trauma (Figure D3.11.9), imaging will doubtlessly support and improve forensic reconstruction by its possibilities of revealing inner trauma findings. In many cases both MRI and CT will be necessary for optimized
© 2009 by Taylor & Francis Group, LLC
FIGURE D3.11.6 A pedestrian’s foot was run over by a car. MRI shows extensive bone marrow edema of the foot skeleton, especially pronounced in the fourth metatarsal bone (A, frame), the navicular bone (arrows in C), and the capita of the other metatarsal bones (A,B). No fractures were present. Besides distinct bone bruising also in the first metatarsal bone (B), the underlying fat tissue presents with hemorrhage that is consistent with abutment injury (B, arrows). The extent of bone marrow edema and the subcutaneous lesions clearly point toward broad blunt force impact and therefore approve the reported trauma mechanism.
results concerning the assessment of soft-tissue, bone, and organ findings. D3.11.4.3 Child Abuse Physical and sexual abuse of children is an important issue in clinical forensic medicine, and the differentiation between accidental and abusive injuries is challenging in many of these cases. All forms of nonaccidental trauma that are known from clinical forensic examinations of adult persons might
370
The Virtopsy Approach
A
B
FIGURE D3.11.7 In living persons, whole-body MRI offers an excellent overview of the injury status following trauma. This 25-year-old male pedestrian, who was intoxicated at the time, was hit by a car. Shortly before, the man had crashed his own car against a wall; the (moderate) impact from that collision was found on the right frontal side of the vehicle. After that, the man walked on a highway, where the second accident occurred. Whole-body T2-weighted MRI reveals distinct hemorrhage of the subcutaneous soft tissues in the left lateral and medial and the right medial knee region (small arrows). Bone marrow edema is easily detected at MRI, as demonstrated in this case (large arrows). A subcutaneous fluid-filled crush cavity that corresponds to the uploading of the man on the vehicle is seen on the lateral side of the left thigh (frame). Altogether, the localization and distribution of the injuries indicate a collision direction from the anterior-left side when the person was hit by the car. As whole-body MRI was performed using a clinical standard sequence in this case, the lower legs were cut. This well demonstrates that for a forensic application it will be important to adapt the clinical sequences to meet the forensic needs.
also occur in children; however, there are some typical patterns that are observed especially in children following abuse. One of these is the characteristic intracranial injury pattern following shaking in shaken baby cases, mostly causing subdural hematoma, which is often combined with other forms
© 2009 by Taylor & Francis Group, LLC
C
FIGURE D3.11.8 This sagittal T2-weighted MR image shows hyperintense edema of the navicular bone (A, frame) that occurred due to fracture (B, axial slice, arrow) that had been invisible in the x-ray examinations. The patient who reported that he had been injured when stumbling on the stairs also suffered bone bruising of the calcaneus (C). In contrast to the findings shown in Figure D3.11.5, where broad blunt trauma had caused the injuries, no subcutaneous hemorrhage was detected in this case, and bone bruising was restricted to the navicular and calcaneus bones, both being consistent with the reported accidental trauma mechanism.
of intracranial hemorrhage and characteristic soft-tissue and bone injuries following violent grabbing of the child [3,37–39]. Other injuries that typically occur following nonaccidental trauma predominantly in small children are some characteristic fracture types (e.g., spiral-oblique diaphyseal fractures of the long bones, or multiple rib fractures without an adequate history) [40]. In sexual abuse cases, the range is from no injuries to abrasions and bruises or other indicators of violence to severe injuries of the genitoanal region and others [41]. Abusive trauma in newborns and small children might remain undetected in external examination of the body due to the often barely visible signs that traumatic impact might cause in this age group. Furthermore, young children have far reduced possibilities to articulate themselves toward others, thus making the forensic assessment even more difficult. Radiological techniques have been applied in suspected cases of child abuse since many years, and meanwhile broad experience exists in this field in clinical radiology. Imaging plays an important and well-established role in today’s evaluation of infant trauma (Figure D3.11.10). Besides classical
Incident-Specific Cases
371
RES/SHADE/SURF
A
C
B O
97 72
W C
507 352
RES/SHADE/SURF
B
B O
97 72
W C
D
507 352
E
FIGURE D3.11.9 The differentiation between accidental and inflicted trauma is of high relevance in clinical forensic medicine. In this case of a 16-year-old boy, forensic expertise had to clarify if the person was beaten with a baseball bat or if he fell from the first floor when climbing on a building at night. Besides the general case circumstances that were used for the forensic reconstruction, imaging gave important clues for the sequence of events. The boy died several hours after the incident, and the body was cremated before forensic examination was ordered. For this reason, forensic expertise was mainly based on the clinical imaging data. As shown in three-dimensional volume-rendered CT (A,B) and the axial CT slices (C), a frontal burst fracture system with only minimal impression was found. The fracture system included the frontal and nasal bones, the sinuses, both orbitae, the right maxilla, and the right frontal and medial cranial fossa (arrows in A,B,C) and indicated severe blunt trauma to the frontal region. The intracranial findings (D) included left-sided epidural hematoma and intracerebral hemorrhage in both frontoparietobasal lobes that were seen as hyperdense areas and corresponded to coup findings. Contrecoup hemorrhage was seen in the left occipital lobe (D, arrows), confirming massive deceleration trauma. Brain swelling and midline shifting were present (D). The conclusion that the injuries occurred due to a fall from height was supported by the evidence of massive lung contusions (E), neck soft-tissue emphysema, and a left-sided elbow fracture.
© 2009 by Taylor & Francis Group, LLC
372
The Virtopsy Approach
A
B
finding is often present for weeks and months after the traumatic impact, MRI may give important clues for the forensic evaluation of earlier trauma. Forensic imaging reveals the internal trauma sequelae in combination with the classical forensic external examination, and case history is likely to improve the medicolegal processing in suspected child abuse cases. A summary of the state of the art in pediatric radiology of abuse is provided by Kleinman [40], and many recent articles are found on the topics of nonaccidental injury assessment in children [37,38,48–57]. These imaging studies were, however, performed in clinical radiology units with a clinical and not a primary forensic indication. In the majority of suspected child abuse cases the CT and MRI scans are currently performed in the absence of forensic experts, who often get the clinical imaging data not before days or even weeks after the incident. As the main focus in clinical medicine is on the injury assessment from a therapeutic and prognostic view, the assessment of the need and forms of treatment is standing in the foreground when the radiological examinations are planned and performed. For this reason, for example, an MRI of the body to look for clinically negligible but forensically highly relevant subcutaneous tissue lesions will not be available today when it is not specially ordered in advance. The establishment of interdisciplinary child protection groups in clinical pediatric departments will help to provide forensic input in an early stage of case evaluation and processing. Medicolegal research has to establish specific imaging indications and protocols to cover not only the clinical but also the forensic needs when children are assessed following abuse.
D3.11.4.4 Abuse of the Elderly FIGURE D3.11.10 This 13-day-old baby was submitted to the hospital after the parents had observed scalp hemorrhage. The first-born child was in good condition, and no trauma history was reported. The family had tried to have a baby for more than 10 years before pregnancy. At conventional radiography and ultrasound, a linear singular fracture line was found in the frontoparietal right skull that was considered as “days old.” MRI showed normal brain structures and did not reveal intracranial pathologies. The fact that at MRI the scalp tissues showed no signs of swelling or acute hemorrhage in the fracture region (A, frame) supported the estimation that birth trauma at vaginal delivery caused the injury. The externally visible scalp hematoma (B, frame) appeared at least several days old.
roentgenography, ultrasonographic examination, and bone scintigraphy, CT and MRI have become the main clinical investigative tools in children. Most neuroradiologic findings, including shearing and rotational injuries or gliding contusions, are accessible by CT and MRI. The skeleton can be assessed not only by conventional radiography and bone scintigraphy but also by MRI depicting signs of minor skeletal trauma. MRI represents the only imaging method that objectively shows the pathology of the bone marrow space, revealing, among others, bone marrow edema as a result of trauma, known as bone bruises [42–47]. As this
© 2009 by Taylor & Francis Group, LLC
Similar to child abuse, abuse of elderly persons is an often underestimated form of incident. Depending on the applied mechanism of violence, more or less discrete findings might be present, mainly lesions of the subcutaneous tissue due to blunt force trauma [58–60]. The differentiation between accidental and inflicted trauma might be difficult, as aged persons are often in a generally reduced physical or cognitive condition. As with small children, the less information can be obtained from the victim, the more importance must be attached to an objective and thorough evaluation of all findings present, including if possible the inner injuries. For the assessment of blunt force injuries, the application of wholebody MRI might be advantageous in future clinical forensic examinations of potentially abused persons. Neck strangulation sequelae will be detected best when performing neck MRI (Section D3.11.4.15). CT excellently reveals bone injuries and can therefore be helpful not only for the clinical but also for the forensic diagnosis of injuries due to, for example, falls, while MRI will depict specific findings such as bone bruises (Section D3.11.4.3). The forensic assessment of other forms of maltreatment of elderly persons such as intoxication or suffocation attempts will, however, hardly be simplified by means of radiology.
Incident-Specific Cases
D3.11.4.5 Torture The sequelae of various forms of torture are sometimes subject to forensic evaluation. In most of these cases, the maltreatment has been applied months to years before the forensic examination is requested. Depending on the applied form of torture and the time span between the incident and the forensic examination, often no visible or only very discrete lesions will be present, mostly in the form of scars in the body surface tissues [61–65]. This explains the limited benefit that must be expected when applying radiological imaging techniques in these cases. If the inflicted torture caused severe injuries such as bone fractures or even amputations, forensic radiological examination even a long time after the incident is presumably well suited to improve the forensic evaluation and interpretation. D3.11.4.6 Self-Inflicted Injury Self-inflicted trauma comprises several issues, each more or less characteristic for the specific incident form [3]. Injuries that are, for example, self-inflicted due to psychiatric disorders predominantly comprise superficial skin lesions due to sharp or semisharp force. In these cases, a forensic radiological examination will provide no benefit, as these often very characteristic lesions are in most cases easily seen upon external examination of the body surface. In contrast to superficial injuries, the self-infliction of foreign materials will in many cases, depending on the materials used, be detected and characterized well using imaging methods [66–68]. Regarding another issue of self-inflicted trauma, which is insurance fraud, radiology can be helpful with the often difficult differentiation of accidental and self-inflicted injury. In the majority of these cases, lesions of the extremities are present (e.g., amputations of digits or toes). As clinical radiological examinations are usually performed, these data will be accessible also for subsequent forensic analysis. It might, however, be of advantage to perform additional forensic radiological examinations to improve the basis for medicolegal interpretation, as MRI could, for example, help in assessing accompanying soft-tissue injuries, and CT allows the compilation of highresolution three-dimensional reconstructions that presumably improve the evaluation of the injury-causing mechanism. Additionally, imaging seems well suited for the estimation of long-time sequelae of injury in living persons, which can be also of specific interest in these cases. D3.11.4.7 Body Packing The value of conventional radiography in cases of body packing has been proven in numerous clinical examinations [69–72]. Depending on the materials used for body packing, the foreign objects are easily detected at imaging in most cases. Besides radiography, CT and MRI have a potential for displaying the materials and their location inside the body. If in suspected body packing cases clinical imaging has not been performed up to the time when the case is reported to
© 2009 by Taylor & Francis Group, LLC
373
the forensic examiner, clinical forensic radiological examination should be considered. D3.11.4.8 Sexual Assault Without doubt, radiological methods will play only a minor role in the future evaluation of sexual assault. Only in cases with accompanying extragenital trauma, clinical forensic MRI, for example, will be beneficial by depicting hemorrhage and edema of the body tissues, which might be helpful at the assessment of the sequence of events or, for example, the life-threatening quality of the accompanying injuries. In the increasing number of cases who report sexual violation without having suffered any assault, radiology might help identify these by providing evidence that excludes the presence of, for example, blunt force trauma and therefore might be strongly contradictory to the reported incident. To date, no clinical experience exists with the application of MRI or CT of the genitoanal region in sexual assault cases. However, if there was suspicion of severe rectal or vaginal injury being associated with pneumoperitoneum, conventional radiography has been applied successfully in the past [73,74]. Reflecting that genital injuries of this severity grade are rare in sexual abuse cases, the value of imaging the genitoanal region for the forensic assessment of sexual abuse remains in question. D3.11.4.9 Personal Fitness: Fitness to Drive, Fitness to Act, Fitness to Be Imprisoned Issues regarding personal acute or long-time fitness are regularly addressed in forensic examinations. The medical record, including radiological examinations, is frequently taken into account for forensic expertise in these cases. Regarding the medicolegal evaluation of the personal fitness, radiological examinations will help with the assessment of relevant findings and the current clinical status of the examined person. In persons who, for example, caused a motor vehicle accident, the application of forensic imaging shortly after the incident might help in identifying clinical conditions that are suited for provoking the acute driving disorder. If clinical radiological data are used for the forensic assessment, it might be important to perform a second-look reading session together with an experienced radiologist to address also the specific forensic needs. D3.11.4.10 Forensic Psychiatry Forensic psychiatry is a very specialized field of forensic medicine that shall not be discussed in detail in this chapter. The application of radiological techniques has a long tradition in the clinical evaluation of acute and chronic psychiatric disorders, and modern imaging offers some possibilities in the living that are also of specific forensic interest. Functional MRI, for example, is expected to provide interesting results when applied to the assessment of criminal offenders [75–80]. A field that is forensically relevant but of high ethical complexity is the usage of imaging methods as a “lie detector.” Functional MRI has been tested for this application in the past years
374
The Virtopsy Approach
[81,82]. It remains a question of further forensic radiological, ethical, and psychiatric research to establish reasonable indications and interpretation guidelines in these specific cases. D3.11.4.11 Civil Law Issues: Claims for Damages Some forensic issues are especially relevant in the context of civil lawsuits. In the course of civil law proceedings, the forensic expert is frequently asked for an evaluation of longtime sequelae following a traumatic event that is of importance regarding claims for damages. Mostly in these cases the clinical radiological data are available for the forensic assessment. If there is, however, a lack of up-to-date imaging examinations or if the clinical data do not sufficiently cover the forensic needs, additional clinical forensic CT and MRI might be indicated. These might offer improved assessment and interpretation of the current and prognostic status of the patient following trauma. D3.11.4.12 Intoxication According to some radiological case reports, CT and especially MRI have shown an ability to discover specific findings following recent or chronic intoxication with various substances [83–97]. According to the clinical literature, the basal ganglia are regularly affected at MRI in substance abuse or acute and chronic intoxication. However, representing relative new techniques, the possibilities of diffusion-weighted MRI and MR spectroscopy give rise to expectations for some forensically interesting applications in the future evaluation of acute and chronic poisoning in living persons. D3.11.4.13 Medical Malpractice For detailed information about forensic imaging following medical malpractice, see Chapter D3.12. D3.11.4.14 Age Estimation The means of forensic age estimation and the possible application of radiological methods are described in detail in Chapter XY.
D3.11.5 THE USE OF CLINICAL IMAGES FOR FORENSIC PURPOSES Imaging of accidental and nonaccidental trauma plays a major role in today’s clinical patient management, and in an increasing number of cases these clinical imaging data also serve as a basis for forensic expertise. Especially in the fields of child abuse and medical malpractice as well as for the determination of the extent and severity of traumatic injuries, the forensic examiners often use clinical radiological data in addition to conventional medicolegal methods. However, in these cases the radiological examinations have been initially performed from a clinical view and therefore do not cover the whole forensic spectrum that could be served by means of radiology (Figure D3.11.7 and Figure D3.11.11). If the forensic
© 2009 by Taylor & Francis Group, LLC
FIGURE D3.11.11 In 2001, the New England Journal of Medicine published this axial CT slice of a patient having suffered craniocerebral trauma. In the figure legends the presence of all four intracerebral hemorrhage types on one CT slice was delineated (black and white arrows). This image well demonstrates the difference between the clinical and forensic approach: The authors did not describe the scalp hematoma that was visible in the left parietal region. This finding is important for forensic case assessment, as it allows determining the impact axis (blue arrow), which is of relevance for the forensic reconstruction of the sequence of events. (From Mattiello, J. A. and Munz, M., New Engl J Med, 344, 580, 2001. With permission.)
expert uses only the written reports of the examining radiologist, he or she might oversee forensically important findings. For example, clinical radiological evaluation is rarely focused on in the detection of superficial soft-tissue lesions, which are of utmost importance for the forensic analysis, and the subcutaneous regions are often cut off to save time at MRI or to reduce radiation exposure at CT. Furthermore, most radiologists have limited forensic knowledge and do not focus their evaluation on specific forensic issues such as the fracture system or, again, soft-tissue lesions. If the forensic expert uses the clinical data as the basis for his or her analysis without a “second-look” image reading session that is focused on forensically relevant findings, a number of clinically irrelevant but possibly important forensic findings might escape from being recognized. In conclusion, the extensive clinical experience with imaging trauma provides an excellent basis for clinical forensic radiology. To obtain optimized results also for forensic case management, it seems important, however, that the medicolegal examiner performs a (secondary) image reading of clinical MRI and CT data in the presence of and with the expertise of an experienced clinical radiologist [98].
Incident-Specific Cases
D3.11.6 THE ROLE OF CLINICAL FORENSIC IMAGING IN THE JURIDICAL CONTEXT It is fundamental to remain conscious of the fact that clinical forensic imaging will always be only part of the forensic evaluation process. The forensic expert must be highly alert to keep from overestimating the value of the radiological images in the case context. The whole case circumstances and all findings obtained always must be taken as the basis for forensic expertise. It is, for example, hardly possible to perform a reliable reconstructive analysis in a pedestrian who was hit by a car without considering the technical information about the damage status of the involved vehicle. Furthermore, as the clinical radiologist is rarely involved in the legal proceeding of a case, the forensic examiner must critically review and analyze the radiologist’s diagnoses and remarks before using these data. When providing forensic expertise in clinical forensic cases, three main issues have to be addressed regularly: (1) the classification of the injury type; (2) the classification of the severity of the injuries (endangerment to life) and their sequelae; and (3) forensic reconstruction issues (determination of the sequence of events). According to clinical experience and the results from the Virtopsy Studies, clinical forensic imaging is especially suited to support the interpretation of injury-severity grading and reconstructive issues. As required in most juridical systems, the objective basis for forensic expertise should be as broad as possible to serve as a sound fundament for later juridical treatment. Therefore, the internal body findings should be obtained as far as possible at the forensic evaluation. Preexisting clinical radiological data should be considered and, if necessary, read for a second time for acquiring all forensically relevant facts (Section D3.11.5). The expert report should contain the radiological image interpretation and, for better understanding, also selected images with the relevant findings highlighted or indicated by arrows. For an optimized benefit from a broad clinical forensic application of radiological imaging techniques, it will be essential that the juridical experts get used to the utilization of radiological data for the forensic case interpretation and that they will be informed about the useful indications and the added value of CT and MRI in the forensic context. Applying radiological imaging methods in clinical forensic cases will doubtlessly be associated with an initial increase of cost. The improved basis for the forensic and therefore also the juridical treatment of the cases will, however, help markedly reduce the legal expenses in many cases as the initially clearer situation carries a potential of reducing the number of secondary instance and civil law proceedings.
D3.11.7 ETHICAL AND JURIDICAL CONSIDERATIONS When applying imaging techniques in living persons for forensic purposes only, some ethical issues must be considered. For example, CT and MRI usually require general anesthesia in small infants. The application of CT is generally associated with radiation exposure. It remains a question
© 2009 by Taylor & Francis Group, LLC
375
of ethical discourse whether in such situations imaging can be performed with a forensic indication only. Reflecting the massive impact that (wrong) conviction has on the involved persons, it will be necessary to find ethical consent in these sensible questions. Depending on the legal situation in a country, accordant adaptations to the law might also be required. Another relevant issue that will have ethical and also juridical implications is the fact that imaging might reveal findings that require further diagnostics and treatment and were formerly unknown to the examined person. This situation occurred several times within the clinical forensic strangulation studies of the Virtopsy Project [5] (Figure D3.11.12). When the indication for the examination is only a forensic one, it remains unclear how these additional results should be handled. It will be important that in these cases the forensic examiner takes responsibility also for the information of the patient and the organization of further medical processing. Not only for this reason should a clinical radiologist always be present at the scans even when the imaging examination is performed for forensic purposes only. Being associated with his or her responsibility for the living patient, the forensic doctor is at some increased risk for being accused of medical malpractice when performing clinical forensic radiological
FIGURE D3.11.12 As an additional and unexpected finding, this 45-year-old woman showed signs of left-sided cerebellar insult (frame) at MRI when the neck was examined following manual strangulation. Follow-up examinations that were planned on the basis of the forensic scan’s result depicted no other intracranial or vessel abnormalities, and the cerebellar finding was judged as months to years old. The aetiology remained unclear, though the woman had repeatedly been strangled and severely beaten by her husband. She could not remember any apparent clinical signs of cerebellar insult.
376
The Virtopsy Approach
examinations. A clinical radiologist being present at the time of the examination to read the images previously and as long as the patient is at the scanning unit will help to minimize such risk: If findings are present that may require prompt intervention (e.g., vessel lesions in strangulation cases, signs of deficits in cerebral perfusion, organ lesions), this will more reliably not be overseen when not only the usually radiologically untrained forensic expert is on site.
D3.11.8 CONCLUSION Today’s standard in the forensic examination of living persons is the external inspection of the body. Apart from being a subjective examination method, this has the disadvantage that internal findings widely escape the forensic evaluation. The clinically well-established radiological methods MRI and CT have the potential to provide an additional and objective basis also for forensic evaluation, offering access to internal injuries in the living. Due to the far improved basis for forensic expertise when referring also to the internal injuries, their implementation into the clinical forensic routine process will lead to an increased legal certainty in the juridical processing. However, forensic imaging has entirely different objectives from clinical radiology and therefore requires different evaluation and interpretation of MRI and CT data. Clinical radiology is targeted at patient diagnosis for assessing therapeutic options, while forensic imaging aims at the reconstruction of the sequence of events and the interpretation of severity of injuries, including the life-threatening quality of an act. Although clinical radiological studies and experience serve as a sound basis also for future forensic examinations, it will be inevitable to perform specific forensic MRI and CT studies to lay the scientific fundamentals for a clinical forensic routine application of the radiological methods.
D3.11.9 REFERENCES 1. Pollak, S., Clinical forensic medicine and its main fields of activity from the foundation of the German Society of Legal Medicine until today, Forensic Sci Int, 144, 269 (2004). 2. Gall, J., Centres of excellence are needed for clinical forensic medicine, BMJ, 320, 650 (2000). 3. Brinkmann B, M. B., [Handbook forensic medicine], Vol. 1, Springer, Berlin (2004). 4. Yen, K., Thali, M. J., Aghayev, E., Jackowski, C., Schweitzer, W., Boesch, C., et al., Strangulation signs: initial correlation of MRI, MSCT, and forensic neck findings, J Magn Reson Imaging, 22, 501 (2005). 5. Yen, K., Vock, P., Christe, A., Scheurer, E., Plattner, T., Schon, C., et al., Clinical forensic radiology in strangulation victims: forensic expertise based on magnetic resonance imaging (MRI) findings, Int J Legal Med, 121, 115 (2007). 6. Brogdon, B. G., Forensic Radiology, CRC Press, Boca Raton, FL (1998). 7. Schmeling, A., Olze, A., Reisinger, W., and Geserick, G., Forensic age diagnostics of living people undergoing criminal proceedings, Forensic Sci Int, 144, 243 (2004).
© 2009 by Taylor & Francis Group, LLC
8. Schumacher, M., Oehmichen, M., Konig, H. G., and Einighammer, H., [Intravital and postmortal CT examinations in cerebral gunshot injuries], Rofo, 139, 58 (1983). 9. Scialpi, M., Boccuzzi, F., Romeo, F., Ax, G., Scapati, C., Rotondo, A., et al., [Computerized tomography in craniocerebral, maxillofacial, cervical, and spinal gunshot wounds. Part II—Clinical contribution and medico-legal aspects], Radiol Med (Torino), 92, 693 (1996). 10. Yen, K., Vock, P., Tiefenthaler, B., Ranner, G., Scheurer, E., Thali, M. J., et al., Virtopsy: forensic traumatology of the subcutaneous fatty tissue; multislice computed tomography (MSCT) and magnetic resonance imaging (MRI) as diagnostic tools, J Forensic Sci, 49, 799 (2004). 11. Dirnhofer, R., Jackowski, C., Vock, P., Potter, K., and Thali, M. J., VIRTOPSY: minimally invasive, imaging-guided virtual autopsy, Radiographics, 26, 1305 (2006). 12. Jackowski, C., Thali, M., Sonnenschein, M., Aghayev, E., Yen, K., Dirnhofer, R., et al., Visualization and quantification of air embolism structure by processing postmortem MSCT data, J Forensic Sci, 49, 1339 (2004). 13. Jackowski, C., Thali, M. J., Buck, U., Aghayev, E., Sonnenschein, M., Yen, K., et al., Noninvasive estimation of organ weights by postmortem magnetic resonance imaging and multislice computed tomography, Invest Radiol, 41, 572 (2006). 14. Thali, M. J., Yen, K., Schweitzer, W., Vock, P., Boesch, C., Ozdoba, C., et al., Virtopsy, a new imaging horizon in forensic pathology: virtual autopsy by postmortem multislice computed tomography (MSCT) and magnetic resonance imaging (MRI)—a feasibility study, J Forensic Sci, 48, 386 (2003). 15. Gniadecka, M. and Danielsen, L., High-frequency ultrasound for torture-inflicted skin lesions, Acta Derm Venereol, 75, 375 (1995). 16. Seifert, D. and Puschel, K., Subgaleal hematoma in child abuse, Forensic Sci Int, 157, 131 (2006). 17. Martino, F., Laforgia, R., Rizzo, A., Dicandia, V., Strada, A., Macarini, L., et al., [The echographic assessment of traumatic rib lesions], Radiol Med (Torino), 94, 166 (1997). 18. Akopov, V. I. and Kuryshev, A. N., [Possibility of using ultrasound in the forensic medical study of bone injuries], Sud Med Ekspert, 23, 20 (1980). 19. Cheng, Y., [Analysis of 58 cases of the usage of B’-scan in clinical forensic medical examination], Fa Yi Xue Za Zhi, 13, 85 (1997). 20. Cotton, J. M., Cooke, J. C., and Monaghan, M. J., Forensic echocardiography: a case in point, Echocardiography, 17, 193 (2000). 21. Mele, A., Maione, S., Giunta, A., Mele, R., and Mele, A. F., Usefulness of monodimensional echocardiography in forensic medicine, Acta Med Leg Soc (Liege), 32, 437 (1982). 22. Reay, D. T. and Holloway, G. A., Jr., Changes in carotid blood flow produced by neck compression, Am J Forensic Med Pathol, 3, 199 (1982). 23. Clarot, F., Vaz, E., Papin, F., and Proust, B., Fatal and nonfatal bilateral delayed carotid artery dissection after manual strangulation, Forensic Sci Int, 149, 143 (2005). 24. Blanco Pampin, J., Morte Tamayo, N., Hinojal Fonseca, R., Payne-James, J. J., and Jerreat, P., Delayed presentation of carotid dissection, cerebral ischemia, and infarction following blunt trauma: two cases, J Clin Forensic Med, 9, 136 (2002). 25. Hierholzer, J., Tantow, H., Cordes, M., Maurer, J., Keske, U., Koch, E., et al., [Roentgen diagnosis of body packers— radiological and forensic considerations], Aktuelle Radiol, 5, 157 (1995).
Incident-Specific Cases
26. Yen, K., Sonnenschein, M., Thali, M. J., Ozdoba, C., Weis, J., Zwygart, K., et al., Postmortem multislice computed tomography and magnetic resonance imaging of odontoid fractures, atlantoaxial distractions and ascending medullary edema, Int J Legal Med, 119, 129 (2005). 27. Aghayev, E., Jackowski, C., Sonnenschein, M., Thali, M., Yen, K., and Dirnhofer, R., Virtopsy hemorrhage of the posterior cricoarytenoid muscle by blunt force to the neck in postmortem multislice computed tomography and magnetic resonance imaging, Am J Forensic Med Pathol, 27, 25 (2006). 28. Aghayev, E., Thali, M., Jackowski, C., Sonnenschein, M., Yen, K., Vock, P., et al., Virtopsy—fatal motor vehicle accident with head injury, J Forensic Sci, 49, 809 (2004). 29. Aghayev, E., Thali, M. J., Sonnenschein, M., Hurlimann, J., Jackowski, C., Kilchoer, T., et al., Fatal steamer accident; blunt force injuries and drowning in post-mortem MSCT and MRI, Forensic Sci Int, 152, 65 (2005). 30. Lupetin, A. R., Hollander, M., and Rao, V. M., CT evaluation of laryngotracheal trauma, Semin Musculoskelet Radiol, 2, 105 (1998). 31. Stanley, R. B., Jr. and Hanson, D. G., Manual strangulation injuries of the larynx, Arch Otolaryngol, 109, 344 (1983). 32. Plattner, T., Bolliger, S., and Zollinger, U., Forensic assessment of survived strangulation, Forensic Sci Int, 153, 202 (2005). 33. Kaale, B. R., Krakenes, J., Albrektsen, G., and Wester, K., Head position and impact direction in whiplash injuries: associations with MRI-verified lesions of ligaments and membranes in the upper cervical spine, J Neurotrauma, 22, 1294 (2005). 34. Krakenes, J. and Kaale, B. R., Magnetic resonance imaging assessment of craniovertebral ligaments and membranes after whiplash trauma, Spine, 31, 2820 (2006). 35. Krakenes, J., Kaale, B. R., Moen, G., Nordli, H., Gilhus, N. E., and Rorvik, J., MRI assessment of the alar ligaments in the late stage of whiplash injury—a study of structural abnormalities and observer agreement, Neuroradiology, 44, 617 (2002). 36. Ovadia, D., Steinberg, E. L., Nissan, M. N., and Dekel, S., Whiplash injury—a retrospective study on patients seeking compensation, Injury, 33, 569 (2002). 37. Vinchon, M., Defoort-Dhellemmes, S., Desurmont, M., and Dhellemmes, P., Accidental and nonaccidental head injuries in infants: a prospective study, J Neurosurg, 102, 380 (2005). 38. Leestma, J. E., Case analysis of brain-injured admittedly shaken infants: 54 cases, 1969–2001, Am J Forensic Med Pathol, 26, 199 (2005). 39. Lonergan, G. J., Baker, A. M., Morey, M. K., and Boos, S. C., From the archives of the AFIP. Child abuse: radiologicpathologic correlation, Radiographics, 23, 811 (2003). 40. Kleinman, P., Diagnostic imaging of child abuse, Mosby, London (1990). 41. Hobbs, C. J. and Osman, J., Genital injuries in boys and abuse, Arch Dis Child, 92, 328 (2007). 42. Boks, S. S., Vroegindeweij, D., Koes, B. W., Hunink, M. G., and Bierma-Zeinstra, S. M., Follow-up of occult bone lesions detected at MR imaging: systematic review, Radiology, 238, 853 (2006). 43. Scheunemann, D., Lehmann, W., Briem, D., Stork, A., Windolf, J., Rueger, J. M., et al., [Clinical relevance of “bone bruise” detected by MRI following spinal injuries in children], Unfallchirurg, 108, 638 (2005).
© 2009 by Taylor & Francis Group, LLC
377
44. Vincken, P. W., Ter Braak, B. P., van Erkel, A. R., Coerkamp, E. G., Mallens, W. M., and Bloem, J. L., Clinical consequences of bone bruise around the knee, Eur Radiol, 16, 97 (2006). 45. Rogers, L. F., To see or not to see, that is the question: MR imaging of acute skeletal trauma, Am J Roentgenol, 176, 1 (2001). 46. Sanders, T. G., Medynski, M. A., Feller, J. F., and Lawhorn, K. W., Bone contusion patterns of the knee at MR imaging: footprint of the mechanism of injury, Radiographics, 20 Spec No, S135 (2000). 47. Lal, N. R., Jamadar, D. A., Doi, K., Newman, J. S., Adler, R. S., Uri, D. S., et al., Evaluation of bone contusions with fat-saturated fast spin-echo proton-density magnetic resonance imaging, Can Assoc Radiol J, 51, 182 (2000). 48. Kemp, A. M., Investigating subdural haemorrhage in infants, Arch Dis Child, 86, 98 (2002). 49. Soto-Ares, G., Denes, M., Noule, N., Vinchon, M., Pruvo, J. P., and Gosset, D., [Subdural hematomas in children: role of cerebral and spinal MRI in the diagnosis of child abuse], J Radiol, 84, 1757 (2003). 50. Stoodley, N., Controversies in non-accidental head injury in infants, Br J Radiol, 79, 550 (2006). 51. Vinchon, M., Noizet, O., Defoort-Dhellemmes, S., SotoAres, G., and Dhellemmes, P., Infantile subdural hematomas due to traffic accidents, Pediatr Neurosurg, 37, 245 (2002). 52. Vinchon, M., Noule, N., Tchofo, P. J., Soto-Ares, G., Fourier, C., and Dhellemmes, P., Imaging of head injuries in infants: temporal correlates and forensic implications for the diagnosis of child abuse, J Neurosurg, 101, 44 (2004). 53. Carty, H. and Pierce, A., Non-accidental injury: a retrospective analysis of a large cohort, Eur Radiol, 12, 2919 (2002). 54. Jaspan, T., Griffiths, P. D., McConachie, N. S., and Punt, J. A., Neuroimaging for non-accidental head injury in childhood: a proposed protocol, Clin Radiol, 58, 44 (2003). 55. Stoodley, N., Neuroimaging in non-accidental head injury: if, when, why and how, Clin Radiol, 60, 22 (2005). 56. Demaerel, P., Casteels, I., and Wilms, G., Cranial imaging in child abuse, Eur Radiol, 12, 849 (2002). 57. Stover, B., [Radiologic diagnosis of the battered child syndrome], Monatsschr Kinderheilkd, 134, 322 (1986). 58. Burgess, A. W. and Clements, P. T., Elder abuse: a call to action for forensic nurses, J Forensic Nurs, 2, 110 (2006). 59. Collins, K. A., Elder maltreatment: a review, Arch Pathol Lab Med, 130, 1290 (2006). 60. Wiglesworth, A., Mosqueda, L., Burnight, K., Younglove, T., and Jeske, D., Findings from an elder abuse forensic center, Gerontologist, 46, 277 (2006). 61. Bose, T. K. and Basu, R., Torture in the eyes of a medicologist, J Indian Med Assoc, 98, 318 (2000). 62. Petersen, H. D. and Wandall, J. H., Evidence of physical torture in a series of children, Forensic Sci Int, 75, 45 (1995). 63. Forrest, D. M., Examination for the late physical after effects of torture, J Clin Forensic Med, 6, 4 (1999). 64. Leth, P. M. and Banner, J., Forensic medical examination of refugees who claim to have been tortured, Am J Forensic Med Pathol, 26, 125 (2005). 65. Mirzaei, S., Knoll, P., and Kohn, H., [Medical aspects of objectifying torture sequels], Wien Klin Wochenschr, 116, 568 (2004). 66. Kantarci, M., Ogul, H., and Karasen, R. M., Detection of a giant wooden foreign body with multidetector computed tomography and multiplanar reconstruction imaging, Am J Emerg Med, 25, 211 (2007).
378
67. Litvack, Z. N., Hunt, M. A., Weinstein, J. S., and West, G. A., Self-inflicted nail-gun injury with 12 cranial penetrations and associated cerebral trauma. Case report and review of the literature, J Neurosurg, 104, 828 (2006). 68. Strub, W. M. and Weiss, K. L., Self-inflicted transorbital and intracranial injury from eyeglasses, Emerg Radiol, 10, 109 (2003). 69. Freislederer, A., Bautz, W., and Schmidt, V., [Body packing: the value of modern imaging procedures in the detection of intracorporeal transport media], Arch Kriminol, 182, 143 (1988). 70. Hergan, K., Kofler, K., and Oser, W., Drug smuggling by body packing: what radiologists should know about it, Eur Radiol, 14, 736 (2004). 71. Wackerle, B., Rupp, N., von Clarmann, M., Kahn, T., Heller, H., and Feuerbach, S., [Detection of narcotic-containing packages in “body-packers” using imaging procedures. Studies in vitro and in vivo], Rofo, 145, 274 (1986). 72. Wehr, K. and Alzen, G., [Perfected, professional bodypacking], Z Rechtsmed, 103, 63 (1989). 73. Manchanda, R. and Refaie, A., Acute pneumoperitoneum following coitus, Cjem, 7, 51 (2005). 74. Gantt, C. B., Jr., Daniel, W. W., and Hallenbeck, G. A., Nonsurgical pneumoperitoneum, Am J Surg, 134, 411 (1977). 75. Meloy, J. R. and Fisher, H., Some thoughts on the neurobiology of stalking, J Forensic Sci, 50, 1472 (2005). 76. Dressing, H., Obergriesser, T., Tost, H., Kaumeier, S., Ruf, M., and Braus, D. F., [Homosexual pedophilia and functional networks—an fMRI case report and literature review], Fortschr Neurol Psychiatr, 69, 539 (2001). 77. Brower, M. C. and Price, B. H., Neuropsychiatry of frontal lobe dysfunction in violent and criminal behaviour: a critical review, J Neurol Neurosurg Psychiatry, 71, 720 (2001). 78. Bufkin, J. L. and Luttrell, V. R., Neuroimaging studies of aggressive and violent behavior: current findings and implications for criminology and criminal justice, Trauma Violence Abus, 6, 176 (2005). 79. Kiehl, K. A., Smith, A. M., Hare, R. D., Mendrek, A., Forster, B. B., Brink, J., et al., Limbic abnormalities in affective processing by criminal psychopaths as revealed by functional magnetic resonance imaging, Biol Psychiatry, 50, 677 (2001). 80. Muller, J. L., Sommer, M., Wagner, V., Lange, K., Taschler, H., Roder, C. H., et al., Abnormalities in emotion processing within cortical and subcortical regions in criminal psychopaths: evidence from a functional magnetic resonance imaging study using pictures with emotional content, Biol Psychiatry, 54, 152 (2003). 81. Appelbaum, P. S., Law & psychiatry: the new lie detectors: neuroscience, deception, and the courts, Psychiatr Serv, 58, 460 (2007). 82. Zhao, H. and Kang, M., [Technical development of detecting deception], Fa Yi Xue Za Zhi, 23, 52 (2007). 83. Atre, A. L., Shinde, P. R., Shinde, S. N., Wadia, R. S., Nanivadekar, A. A., Vaid, S. J., et al., Pre- and posttreatment MR imaging findings in lead encephalopathy, Am J Neuroradiol, 27, 902 (2006). 84. Hantson, P. and Duprez, T., The value of morphological neuroimaging after acute exposure to toxic substances, Toxicol Rev, 25, 87 (2006). 85. Hopkins, R. O., Fearing, M. A., Weaver, L. K., and Foley, J. F., Basal ganglia lesions following carbon monoxide poisoning, Brain Inj, 20, 273 (2006).
© 2009 by Taylor & Francis Group, LLC
The Virtopsy Approach
86. Kondo, A., Saito, Y., Seki, A., Sugiura, C., Maegaki, Y., Nakayama, Y., et al., Delayed neuropsychiatric syndrome in a child following carbon monoxide poisoning, Brain Dev, 29, 174 (2007). 87. Molloy, S., Soh, C., and Williams, T. L., Reversible delayed posthypoxic leukoencephalopathy, Am J Neuroradiol, 27, 1763 (2006). 88. Spampinato, M. V., Castillo, M., Rojas, R., Palacios, E., Frascheri, L., and Descartes, F., Magnetic resonance imaging findings in substance abuse: alcohol and alcoholism and syndromes associated with alcohol abuse, Top Magn Reson Imaging, 16, 223 (2005). 89. Stewart, W. F., Schwartz, B. S., Davatzikos, C., Shen, D., Liu, D., Wu, X., et al., Past adult lead exposure is linked to neurodegeneration measured by brain MRI, Neurology, 66, 1476 (2006). 90. Takao, H., Doi, I., and Watanabe, T., Serial diffusionweighted magnetic resonance imaging in methanol intoxication, J Comput Assist Tomogr, 30, 742 (2006). 91. Tsai, Y. T., Huang, C. C., Kuo, H. C., Wang, H. M., Shen, W. S., Shih, T. S., et al., Central nervous system effects in acute thallium poisoning, Neurotoxicology, 27, 291 (2006). 92. Yamasue, H., Abe, O., Kasai, K., Suga, M., Iwanami, A., Yamada, H., et al., Human brain structural change related to acute single exposure to sarin, Ann Neurol, 61, 37 (2007). 93. Blanco, M., Casado, R., Vazquez, F., and Pumar, J. M., CT and MR imaging findings in methanol intoxication, Am J Neuroradiol, 27, 452 (2006). 94. Caparros-Lefebvre, D., Policard, J., Sengler, C., Benabdallah, E., Colombani, S., and Rigal, M., Bipallidal haemorrhage after ethylene glycol intoxication, Neuroradiology, 47, 105 (2005). 95. Finelli, P. F. and DiMario, F. J., Jr., Hemorrhagic infarction in white matter following acute carbon monoxide poisoning, Neurology, 63, 1102 (2004). 96. Sefidbakht, S., Rasekhi, A. R., Kamali, K., Borhani Haghighi, A., Salooti, A., Meshksar, A., et al., Methanol poisoning: acute MR and CT findings in nine patients, Neuroradiology (2007). 97. Fernandes, J. L., Rocha, A. A., Soares, M. V., and Viana, S. L., Lead arthropathy: radiographic, CT and MRI findings, Skeletal Radiol (2007). 98. Bauer, M., Polzin, S., and Patzelt, D., The use of clinical CCT images in the forensic examination of closed head injuries, J Clin Forensic Med, 11, 65 (2004). 99. Mattiello, J. A. and Munz, M., Images in clinical medicine. Four types of acute post-traumatic intracranial hemorrhage, New Engl J Med, 344, 580 (2001).
D3.12 MEDICAL MALPRACTICE K. Yen, H. Schumacher, and R. Dirnhofer
D3.12.1 INTRODUCTION In 1999, the publication To Err Is Human from the U.S. Institute of Medicine (IOM) of the National Academy of Sciences gained tremendous public attention [1]. In this study, the authors reported extremely large numbers of iatrogenic injuries in the United States, with 44,000 to 98,000 persons being killed per year in U.S. hospitals due to medical malpractice.
Incident-Specific Cases
Although these figures must be seen critically [2], medical error is an important health issue that remains in the focus of the media as well as the public. The expert commissions and arbitration boards of the German Medical Association (Gutachterkommissionen und Schlichtungsstellen bei den Ärztekammern) report a number of approximately 12,000 proven malpractice cases per year in Germany [3], of which in 2006 most were related to skeletal and joint disorders and breast cancer. The most frequently involved hospital-based disciplines in this survey were emergency surgery, general surgery, orthopedics, and gynecology, followed by anesthesia and intensive care medicine. The primarily claimed errors concerned operative therapy, followed by radiological diagnostics. All these data are based on the 2006 Medical Error Reporting System (MERS) data [4] and do not include incourt treated malpractice cases. The IOM defines medical error as “the failure to complete a planned action as intended or the use of a wrong plan to achieve an aim”; an adverse effect is described as “an injury caused by medical management rather than by the underlying disease or condition of the patient” [5]. According to Brennan et al. [6], adverse event injuries are caused by medical management and prolong hospital admission or produce disability at the time of discharge. Virchow [7] defined medical malpractice as “a violation of the generally accepted rules of patient management due to a lack of proper attention or carefulness.” Despite the absence of a worldwide uniform definition of the terms associated with medical malpractice and patient safety, the verification that the patient suffered an injury must always be confirmed before a malpractice claim can be carried out. The preventability of adverse effects represents another important fact that substantially contributes to the legal outcome in these cases. Regarding the manifold possibilities of medical errors and adverse events, no general classification exists to date. Reason’s model [8] separates latent from active failures committed by people who are in direct contact with the patient or system, the latter including “unsafe acts-errors” such as cognitive failures, slips, and procedural violations. Latent errors are defined as errors in the design, organization, training, or maintenance that lead to operator errors. Among these are issues such as time pressure, understaffing, inexperience, and inadequate equipment [8]. The ICECI classification by the World Health Organization [9] differentiates several complications of health care, among those “foreign object left in the body during surgical/medical care,” “adverse incidents associated with medical devices in diagnostic/therapeutic use,” or “unintentional cut, puncture, perforation during surgical/medical care.” In a publication by Weingart et al. [10], who summarized several previous studies, the adverse events that occurred in inpatient care mainly resulted from surgery. The most common nonoperative events were complications from drug treatment, therapeutic mishaps, and diagnostic errors [10]. In his study of 448 malpractice claims, Mallach [11] also found most adverse events in surgery, followed by general and internal medicine,
© 2009 by Taylor & Francis Group, LLC
379
anesthesia, and gynecology. Well known from the media, negligent actions resulting in wrong patient identification, wrong-side surgery, or adverse medical device effects are rather rare but highly noticed events. Far more frequent, however, are misdiagnoses, adverse drug effects, or nosocomial infections that may cause severe to lethal consequences. Even though the experience with postmortem magnetic resonance imaging (MRI) and computed tomography (CT) imaging of malpractice cases to date is only preliminary, the first examined cases point toward a beneficial potential of the radiological methods regarding the assessment of some adverse effects. Especially the question if a person suffered (physical) injury could be addressed in an improved manner in a number of cases. This chapter describes the first experiences with forensic imaging of medical malpractice and the potential indications of CT and MRI in future forensic expert testimony.
D3.12.2 THE MEDICAL MALPRACTICE CLAIM In the German-speaking area, there are principally two ways of treating medical malpractice claims. First, the expert commissions and arbitration boards of the medical associations offer out-of-court proceedings and expert analysis. At any time, the process can be stopped by the plaintiff and brought into court. The second possibility is in-court enforcement. Court proceedings of medical malpractice cases follow the usual legal praxis according to each country’s legal system. For successful medical malpractice claims, all of the following conditions must be fulfilled and proven by the plaintiff [12,13]: 1. The defendant owed a duty of care to the plaintiff. 2. This duty was breached by failing to adhere to the standard of care expected. 3. A damage or injury occurred. 4. Damage or injury occurred due to the violated duty (causality). D3.12.2.1 Expert Testimony in Malpractice Cases Accordingly, in malpractice lawsuits, the following conditions have to be evaluated by the legal and forensic experts from an ex ante view [14]: 1. Assessment of damage (e.g., injury, death): Has the patient suffered injury or damage? 2. Is direct causality (i.e., natural causal connection, condition sine qua non: exclusion that the damage or injury could also have occurred in case of medical action according to medical custom) evident? In other words, as cause of a damage is regarded, each condition cannot be neglected without also neglecting the result (condition sine qua non formula) [15,16].
380
The Virtopsy Approach
3. Is the damage or injury a consequence of mistaken conduct (e.g., error in diagnostics, treatment) or a complication of the malignancy or treatment (diagnostic, therapeutic) itself? In case of mistaken conduct, was the error due to active failure such as procedural violation, slip, lapse, or due to latent and organizational conditions, such as errors in communication or control [8]? 4. Under the doctrine of adequacy (Adäquanztheorie), causation is to be affirmed if an act or omission was adequate to significantly contribute to the occurrence of the damage [17,18]. So the question is if A
C
the act or omission was apt for causing the damage or injury (adequacy of the causal connection). 5. Assessment of the objective foreseeability of the consequences of the unsafe act, with medical custom used as the reference standard (i.e., the quality of care that would be expected of a reasonable practitioner in similar circumstances) [12]. The liability of the wrongdoer exists for all consequences that have to be anticipated abstractly [19,20]. 6. In case of proven objective negligence, the final issue that will be assessed in each individual case is blame, if the duties were accomplishable. For B
D
E
FIGURE D3.12.1 83-year-old female who suffered pulmonary embolism of polymethylmethacrylate (PMMA) during percutaneous vertebroplasty. The procedure was performed minimally invasive and under local anesthesia with conscious sedation. During injection of PMMA into the lumbar vertebrae I–IV the patient became restless and moaned, which was interpreted as pain reaction by the attending doctors. Consecutively, an opiate was given. A decline of oxygen saturation to 92% and an increase of blood pressure to 160/65 mm Hg were observed. After termination of the surgical treatment, the patient was no more responsive to painful and verbal stimuli. Under intensive care the woman survived for 9 days without recovering from loss of consciousness and cardiopulmonary instability. At clinical CT examination 6 h after surgery, high-contrast strings of PMMA were seen in the pulmonary arteries of the right upper, middle, and lower lobes. One PMMA string was found reaching from the body of the fourth vertebra into the inferior vena cava. At echocardiography, the right ventricle appeared massively dilated: (A) The 3D volume-rendered postmortem CT image displays the PMMA string that leaked out from the fourth lumbar vertebra and ascended into the inferior vena cava (arrows). Due to the fact that PMMA material is initially soft and hardens within a few minutes, the string followed the blood flow in the inferior caval vein. (B) Autopsy correlate to (A). CT imaging offers an overview that is superior to autopsy regarding the detection and localization of foreign materials, as demonstrated in this case. At autopsy, the PMMA string broke, although the operation situs was examined with special caution. (C) In this maximum-intensity projection CT image, the distribution of PMMA strings in the right and left pulmonary arteries is well seen (frames). The blue arrows point toward PMMA material in the thoracic vertebrae that was inserted by (uneventful) percutaneous vertebroplasty some years earlier. The first lumbar vertebra after present PMMA insertion (green arrow). (D) 3D air structure reconstruction from the postmortem CT data. The left lung shows reduced ventilation in the lower areas. Again, this finding would be hardly detected at autopsy in the same overview and quality of that offered by CT. (E) At autopsy, PMMA strings were found in the pulmonary arteries, almost completely filling the vessel lumina. (From Stricker K, Orler R, Yen K, Takala J, Luginbuhl M, Anesth Analg 2004; 98:1184–86. With permission.)
© 2009 by Taylor & Francis Group, LLC
Incident-Specific Cases
instance, the question of whether the practitioner asked for the patient’s penicillin allergy and whether he was able to foresee the risk involved needs to be proven (subjective foreseeability). Conditions such as inadequate equipment, understaffing, time pressure, fatigue, or inexperience are also taken into account. If blame (subjective Sorgfaltspflichtverletzung) is given, not only civil but also tortious liability is confirmed. If all these conditions are given, negligence and consecutive civil liability are objectively indicated (objektive Sorgfaltspflichtverletzung). According to a recent report from Studdert et al. [21], in 3% of 1,452 closed U.S. malpractice claims examined no medical injuries were verifiable, and in 37% no error was involved. Claims with no associated injuries or errors were also compensated by the insurance companies in a number of cases (0.4% and 10%, respectively)—however, with the payments being significantly lower than in the other cases. The results of a study by Brennan et al. [22] demonstrated that the severity of the patient’s disability was the most indicative factor for payment to the plaintiff. The direct correlation between the severity of injuries and the financial outcome for the plaintiffs underlines the relevance of the expert witness in these trials.
381
excellent insight into the location and presence of foreign bodies, CT allows their detection and documentation [25], even superior to autopsy (Figure D3.12.1, Figure D3.12.2, and Figure D3.12.3). The possibility of displaying gas and gas distribution in the body tissues is another specific advantage of CT that facilitates the diagnosis in malpractice cases where the question of gas embolism arises (Figure D3.12.4, Figure D3.12.5, and Figure D3.12.6) [26–28]. Furthermore, CT has proven to be helpful with the assessment of organ lesions (e.g., brain injuries [Figure D3.12.3], cardiac diseases) and extravascular blood following trauma or surgery [29–38]. The digital reformatting possibilities of CT are of general forensic interest and serve as an additional and practicable tool that improves the medicolegal evaluation (Figure D3.12.1, Figure D3.12.2, Figure D3.12.3, Figure D3.12.4, Figure D3.12.5, Figure D3.12.6, and Figure D3.12.7). MRI has demonstrated excellent soft-tissue and organ contrast in previous postmortem examinations (see Chapter C4 and Chapter D3). The ability of MRI to depict even small softtissue and organ lesions is of specific interest regarding the assessment of the postsurgical or postinjury status of a patient [39–42]. The unique potential of MRI and its special applications—diffusion-weighted imaging (DWI) and diffusion tensor imaging (DTI)—regarding the depiction of hypoxic lesions
D3.12.2.2 Forensic Assessment in Medical Malpractice Cases When reflecting the results of the aforementioned study [21] and the criteria for the assessment of medical malpractice, it becomes obvious that the evaluation of the presence and nature of damage or injury is a key factor in the forensic evaluation process. If a damage or injury is negated, the malpractice lawsuit will be unsuccessful in most cases. For forensic evaluation in cases resulting in the death of a patient, necropsy is the method used for the diagnosis [10,23]; however, the role of the autopsy regarding quality control and patient safety in hospitals is declining as the numbers of (pathologic) autopsies have significantly and constantly decreased in the past years [24]. Facing this worldwide development, imaging might become an increasingly important tool in the assessment of medical malpractice. D3.12.2.3 Postmortem CT and MRI in Malpractice Cases: Preliminary Experience The general benefits of the application of postmortem imaging are well outlined in this book. Regarding forensic imaging of medical malpractice cases, the experiences from the Virtopsy Project are preliminary, as few cases have been examined. However, postmortem applied CT and MRI have demonstrated that they offer possibilities that are in some aspects superior to necropsy, specifically regarding the assessment of the presence and nature of sustained injury. By offering
© 2009 by Taylor & Francis Group, LLC
FIGURE D3.12.2 80-year-old male who fell down the stairs and suffered an odontoid fracture. The type II fracture was treated surgically two days after the accident by anterior screw fixation. The patient suffered postoperative cardiac arrest and could initially be resuscitated; death occurred 2 days later due to hypoxic encephalopathy. Preexisting cardiopathy with cardiac muscle fibrosis, hypertrophy, and coronary sclerosis had caused cardiovascular failure. The displayed postmortem 2D reformatted CT shows the correct position of the screw and excludes malposition of the odontoid process in a region that is difficult to access at autopsy. This information helped to exclude an adverse event and medical malpractice in this case. (From Yen K, Sonnenschein M, Thali MJ, Ozdoba C, Weis J, Zwygart K, et al., Int J Legal Med 2005;119:129–36. With permission.)
382
The Virtopsy Approach
A
H
CVC
LAO/RAO 8 CRAN/CAUD 7
PAC
B
ET
CERC CERC CERC
PAC PAC
ET
MV
GT
DR
F
PCE
GT
C
D
FIGURE D3.12.3 This 58-year-old male received open heart surgery (mitral valve replacement). During the medical process when the extracorporal circulation was inserted, arterial air embolism occurred that caused severe cerebral hypoxia with consecutively increased brain pressure. A few hours after finishing the operation, the patient died from central respiratory arrest. The reason for the occurrence of arterial air embolism was medical error: The extracorporal circulation unit had been prepared in a false way, with a tube being mounted in the wrong direction and thus not aspirating blood but pumping air into the systemic arterial system: (A) The postmortem 3D reconstructed CT provides a good overview of inserted iatrogen materials and in this case their correct positioning (CVC, central venous catheter; PAC, pulmonary artery/Swan-Ganz catheter; ET, endotracheal tube; CERC, sternum cerclage; MV, replaced mitral valve; DR, drainage tube; GT, gastric tube; PCE, precardial electrodes). Parts of the sternum were removed digitally for better visibility. (B) The 2D reformatted CT slices are important for defining the exact position of foreign materials inside the body. As shown here, the correct placement of the gastric and endotracheal tubes and the central venous catheter is confirmed. (C) Postmortem axial CT displaying signs of massively increased brain pressure (e.g., flattening of the gyri, loss of gray–white matter differentiation, compression of the ventricles). (D) Autopsy correlate to (C) showing massive swelling of the brain with an approximate weight of 1700 g, flattened gyri, and leveled sulci.
in the brain tissue might improve the forensic assessment in malpractice cases associated with cerebral hypoxia (e.g., birth trauma, adverse events in anesthesia, fat embolism) [43–46]. D3.12.2.4 The Use of Radiological Data for Clinical Forensic Case Assessment In living patients, the verification and severity of adverse events sequelae might also be improved by means of clinical forensic CT and MRI, which offer similar advantages when applied postmortem. However, the extensive clinical radiological experiences with the application of CT and MRI are a clear advantage in the examination of living
© 2009 by Taylor & Francis Group, LLC
persons, as the standard imaging protocols do not have to be adapted to postmortem situations, and the radiologists are used to clinical image reading. For instance, the estimation of long-time sequelae of medical actions, which is often a relevant issue in malpractice evaluation, will presumably be more exact in many cases when forensic radiology is applied. In many claimed malpractice cases, clinical imaging data are available also for the forensic analysis. As in other clinical forensic issues, it seems important that these preexisting clinical CT and MRI data are second-read together with an experienced radiologist to obtain optimized forensic radiological results (see Chapter D3.11). Of course, the evaluating
Incident-Specific Cases
A
C
E
383
B
G
D
F
FIGURE D3.12.4 80-year-old female with chronic renal failure. For improved dialysis therapy it was planned to place a continuous venous catheter into the jugular vein. During this procedure, the surgeon recognized a slurping sound with a subsequent increase and later drop of the heart rate. The complication prompted the attending physician to punctuate the left thorax cavity to exclude pneumothorax. At that time the patient was no larger responsive to verbal and painful stimuli, and death was confirmed shortly after. Forensic evaluation of the case revealed that the medical staff members had forgotten to occlude and deaerate the tube before it was inserted in the jugular vein. At the time of catheter introduction, air was sucked to the heart and distributed in the systemic arterial system through an open foramen ovale. According to the autopsy result, death occurred due to right heart and cerebral failure. (A) On the postmortem axial CT slice, air inclusions in the cerebral vessels are easily detected (arrows). They appear as hypodense roundish or elongated structures. (B) At autopsy, air in the deeper brain tissues is usually overseen. The gaseous bubbles in the superficial vessels can hardly be differentiated from air inclusions that occurred at the opening of the cranial cavity. (C) Sagittal reformatted postmortem CT offering excellent demonstration of the air embolism sequelae in the brain. (D) Corresponding 3D air structure reconstruction from the postmortem CT data displaying the distribution of air inside the frontal brain vessels. (E) The axial CT image depicts the distribution of air in the heart, predominantly in the right atrium and ventricle (white arrow) and the coronary arteries. There is also left-sided pneumothorax (orange arrows). (F) Air bubbles in the coronary arteries at autopsy. (G) The sagittal postmortem CT image provides an overview of the air distribution in the body tissues. The right heart (white arrow) and the thorax (orange arrow) are filled with gas, which is also seen in the soft tissues near the puncture region and in some vertebral vessels. At the puncture site, iatrogenic foreign material is visible (blue arrow).
radiologist should not be someone who is directly involved or accused in the malpractice claim. Imaging not only serves as a tool to assess the characteristics and severity of tissue and organ lesions; it can also be used for the exclusion of relevant injuries. Especially in living persons where the radiological methods are the almost only possibility of obtaining inner-body findings, this might be of relevance. If no damage or injury is evident, an expensive malpractice lawsuit can be terminated at an early stage. In postsurgery situations the exclusion of adverse
© 2009 by Taylor & Francis Group, LLC
event injuries and the confirmation that the intervention was performed lege artis are generally of great importance (Figure D3.12.2). D3.12.2.5 Limitations When reflecting the different steps of forensic malpractice assessment (see Section D3.12.2), it becomes obvious that imaging mainly supports the assessment of damage. Therefore, the (forensic) application of imaging is obviously
384
The Virtopsy Approach
A
B
A
C
B
D
C
FIGURE D3.12.5 49-year-old male with tachycardia who underwent electro-ablation of the left atrium. Four weeks after the procedure, the patient’s condition worsened. Clinical CT revealed air bubbles dorsal of the left ventricle near the inferior vena cava, a finding that remained constant for two days and was consistent with the result from endoscopy, which had depicted perforation of the esophagus. No leaking of contrast medium from the heart cavities was observed. Diffusion-weighted MRI showed hyperintense spots in the brain tissue and the cerebellum, which was interpreted as a sign of an ongoing septic-embolic process. When clinical signs of brain death occurred, clinical cranial CT showed signs of massive brain swelling with transtentorial herniation and compression of the ventricles. In the left cerebellum, secondary hemorrhage was seen. Intraparenchymal air inclusions were present in both hemispheres, the basal arteries, and the carotids. At postmortem whole-body CT air inclusions were seen in the heart, brain, and liver tissues. The reason for the fatal outcome was found at autopsy: Coagulation necrosis at the ablation site had caused a fistula between the heart and the esophagus: (A) Gas distribution in the brain parenchyma at postmortem axial CT. The vessels appear partially gas filled, and bubble-like gas amounts are seen predominantly in the white-matter regions. (B) 3D air structure reconstruction from the postmortem CT data. The irregular distribution of gas that is found mostly in the white matter is shown. (C) The liver presents with massive air filling of the blood vessels, producing sort of a “contrast angiogram” at postmortem 3D air structure CT. Gas is also seen in the heart (frame).
© 2009 by Taylor & Francis Group, LLC
FIGURE D3.12.6 A 3 3/4-year-old girl who suffered from acute respiratory infection for 3 days. She was admitted to the hospital and released after 3 hours of uneventful medical supervision. The following day at home, her respiratory situation suddenly worsened, and respiratory arrest occurred within minutes. The child was admitted to the intensive care unit, where resuscitation attempts were unsuccessful: (A) At autopsy, the laryngeal mucosa appeared with reddish discoloration and edema. The larynx did not seem obstructed. (B) The larynx specimen after formalin fixation also showed an open lumen. (C) In contrast to the autopsy findings, postmortem CT clearly revealed an occlusion of the larynx that had caused respiratory failure. In this 3D air structure reconstruction, the occlusion site is well seen (arrow). (D) The axial CT slice that was produced shortly after death also depicts the complete occlusion of the glottis.
useless in cases where no physical lesions are to be expected (e.g., many cases of medication errors, cases with psychiatric sequelae or pain without visible correlates). In other cases, such as wrong patient identification or wrong-side surgery, the error is often obvious; applying additional imaging is not necessary in these. In all cases where MRI and CT are used for the forensic evaluation, the assessor must be aware of the fact that imaging has its specific technical limitations that might reduce the value of the examination (see Chapter B2). Limited spatial resolution, for example, hinders the detection of very small lesions, and artifacts can cause major problems at the image reading. As a general recommendation, the imaging results should always be seen in close addition to all other case information. Last but not least, it is important to reflect that a relevant number of maltreatment claims concern the clinical radiological assessment itself [4,47,48]. Diagnostic errors are among the most frequently claimed malpractice cases, and in many
Incident-Specific Cases
A
385
B
C
FIGURE D3.12.7 56-year-old female who received spondylodesis of the third and fourth lumbar vertebrae. At surgical treatment, two pedicle screws were inserted into the fourth lumbar vertebra. During the operation the cardiopulmonary situation of the patient suddenly deteriorated, and resuscitation measures were applied. Approximately 2 hours after the onset of the complications, the ineffective resuscitation attempts were terminated. Forensic autopsy revealed signs of fatal hemorrhage. The adverse event occurred due to an iatrogenic injury of the inferior vena cava that was severely lacerated at the level of the third and fourth lumbar vertebra. In contrast to the left screw that was in a correct position, the right screw protruded the fourth vertebra in a length of 12.5 cm and caused the vessel injury: (A) 3D reconstruction from the postmortem CT data. The incorrect positioning of the right pedicle screw is clearly demonstrated. (B) Axial minimal intensity projection of the fourth lumbar vertebra displaying the operation site with the false localization of the right screw. (C) Corresponding autopsy specimen. Due to the current lack of routinely applied postmortem angiography techniques, autopsy was in this case superior to imaging regarding the detection and display of the vena cava lesion. The position of the screws was, however, seen better at CT, as demonstrated in this figure.
of these cases diagnostics were mainly based on imaging methods. The forensic expert should keep this in mind when using radiological data as a basis for his or her testimony and should therefore seek the help of experienced clinical radiologists and analyze all imaging data critically and thoroughly. When a case of misdiagnosis must be assessed, the forensic expert should be conscious of the fact that his or her appraisal must be performed from an ex ante view. Secondary postmortem or clinical forensic MRI or CT examinations that were applied using optimized protocols (e.g., whole-body CT offering specific diagnostic possibilities) therefore cannot be used to serve as a direct comparison to what can be expected from routine clinical imaging.
D3.12.3 CONCLUSION Expert testimony is of great importance in court proceedings as well as in the out-of-court treatment of malpractice claims, as the expert reports serve as the main basis for the final decision in most cases. The quality criteria for forensic expertise are high [49] (see Chapter A6), and in addition to the approved qualification of the assessor, the quality of the examination itself should meet the high requirements in this sensitive field. The assessor is strongly recommended to use all case information that is obtainable as the basis for his or her testimony, including also the clinical radiological data. If necessary, specific clinical or postmortem forensic imaging should be applied and evaluated. The analysis, if the plaintiff suffered a verifiable damage or injury, which is substantial for court or out-of-court case treatment, will be greatly improved when being based on additional forensic
© 2009 by Taylor & Francis Group, LLC
radiological methods. However, the forensic assessor must keep the specific limitations of imaging in mind when using CT or MRI as a basis for his or her expertise.
D3.12.4 ACKNOWLEDGMENTS Many thanks go to the Virtopsy Team from the Institute of Forensic Medicine, Bern, for their fundamental support. The authors express their thanks also to the members of the Clinical Radiology Department and the Department of MR Spectroscopy and Methodology of the Insel Hospital Bern for excellent data acquisition and image reading.
D3.12.5 REFERENCES 1. Kohn L, Corrigan J, Donaldson M, To Err Is Human: Building a Safer Health System. Washington, DC, National Academy Press, 1999. 2. Hayward RA, Hofer TP, Estimating hospital deaths due to medical errors: preventability is in the eye of the reviewer. J Am Med Assoc 2001;286:415–20. 3. http://www.bundesaerztekammer.de/page.asp?his= 2.59.5301. 4. http://www.bundesaerztekammer.de/downloads/Gutachterkommissionen_ Statistik_2006.pdf. 5. http://www.iom.edu. 6. Brennan TA, Leape LL, Laird NM, Hebert L, Localio AR, Lawthers AG, et al., Incidence of adverse events and negligence in hospitalized patients. Results of the Harvard Medical Practice Study I. New Engl J Med 1991;324:370–76. 7. Virchow R, Kunstfehler der Ärzte. Aktenstücke des Reichstags des Norddeutschen Bundes. 1870;Anlage 3 zu Nr. 5:7–15. 8. Reason J, Human error: models and management. BMJ 2000;320:768–70.
386
9. http://www.who.int/classifications/icd/adaptations/iceci/ en/. 10. Weingart SN, Wilson RM, Gibberd RW, Harrison B, Epidemiology of medical error. BMJ 2000;320:774–77. 11. Mallach HJ, [Medical malpractice from the viewpoint of legal medicine]. Lebensversicher Med 1986;38:2–11. 12. Studdert DM, Mello MM, Brennan TA, Medical malpractice. New Engl J Med 2004;350:283–92. 13. Keeton W, Dobbs D, Keeton R, Owens D, Prosser and Keeton on the Law of Torts. St. Paul, MN, West Publishing, 1984. 14. Dirnhofer R, Wyler D, [Treatment failure from the viewpoint of forensic medicine]. Ther Umsch 1997;54:272–79. 15. Karner E: § 1295 Nr 3. In: Koziol H, Bydlinski P, Bollenberger R (eds): Kurzkommentar zum ABGB. Vienna, 2005. 16. Schwimann M, Harrer F: § 1295 Nr 6. In: ABGB Praxiskommentar (3) VI. Vienna, 2006. 17. OGH [Austrian Supreme Court] 07.09.1988, 1 Ob 656/88. In: Juristische Blätter 175. Vienna, 1989. 18. Harrer F, Schwimann M: § 1295 Nr 8. In: Praxiskommentar (3) VI. Vienna, 2006. 19. OGH [Austrian Supreme Court] 15.03.2005, 5 Ob 38/05g. In: Recht der Wirtschaft 545, 483. Vienna, 2005. 20. OGH [Austrian Supreme Court] 14.06.2005, 2 Ob 47/05h. In: Ecolex 395, 839, 9. Vienna, 2005. 21. Studdert DM, Mello MM, Gawande AA, Gandhi TK, Kachalia A, Yoon C, et al., Claims, errors, and compensation payments in medical malpractice litigation. New Engl J Med 2006;354:2024–33. 22. Brennan TA, Sox CM, Burstin HR, Relation between negligent adverse events and the outcomes of medical-malpractice litigation. New Engl J Med 1996;335: 1963–67. 23. Garcia-Martin M, Lardelli-Claret P, Bueno-Cavanillas A, Luna-del-Castillo JD, Espigares-Garcia M, Galvez-Vargas R, Proportion of hospital deaths associated with adverse events. J Clin Epidemiol 1997;50:1319–26. 24. Lundberg GD, Low-tech autopsies in the era of high-tech medicine: continued value for quality assurance and patient safety. J Am Med Assoc 1998;280:1273–74. 25. Wieder HA, Feussner H, Rummeny EJ, Gaa J, [Radiological diagnostics for iatrogenic retained foreign bodies after surgery]. Chirurg 2007;78:22–27. 26. Hillewig E, Aghayev E, Jackowski C, Christe A, Plattner T, Thali MJ, Gas embolism following intraosseous medication application proven by post-mortem multislice computed tomography and autopsy. Resuscitation 2007;72: 149–53. 27. Jackowski C, Sonnenschein M, Thali MJ, Aghayev E, Yen K, Dirnhofer R, et al., Intrahepatic gas at postmortem computed tomography: forensic experience as a potential guide for in vivo trauma imaging. J Trauma 2007;62: 979–88. 28. Jackowski C, Thali M, Sonnenschein M, Aghayev E, Yen K, Dirnhofer R, et al., Visualization and quantification of air embolism structure by processing postmortem MSCT data. J Forensic Sci 2004;49:1339–42. 29. Arai A, Shiotani S, Yamazaki K, Nagata C, Kikuchi K, Suzuki M, et al., Postmortem computed tomographic (PMCT) and postmortem magnetic resonance imaging (PMMRI) demonstration of fatal massive retroperitoneal hemorrhage caused by abdominal aortic aneurysm (AAA) rupture. Radiat Med 2006;24:147–49.
© 2009 by Taylor & Francis Group, LLC
The Virtopsy Approach
30. Oyake Y, Aoki T, Shiotani S, Kohno M, Ohashi N, Akutsu H, et al., Postmortem computed tomography for detecting causes of sudden death in infants and children: retrospective review of cases. Radiat Med 2006;24:493–502. 31. Shiotani S, Kohno M, Ohashi N, Yamazaki K, Nakayama H, Watanabe K, et al., Non-traumatic postmortem computed tomographic (PMCT) findings of the lung. Forensic Sci Int 2004;139:39–48. 32. Shiotani S, Watanabe K, Kohno M, Ohashi N, Yamazaki K, Nakayama H, Postmortem computed tomographic (PMCT) findings of pericardial effusion due to acute aortic dissection. Radiat Med 2004;22:405–07. 33. Yamazaki K, Shiotani S, Ohashi N, Doi M, Kikuchi K, Nagata C, et al., Comparison between computed tomography (CT) and autopsy findings in cases of abdominal injury and disease. Forensic Sci Int 2006;162:163–66. 34. Aghayev E, Sonnenschein M, Jackowski C, Thali M, Buck U, Yen K, et al., Postmortem radiology of fatal hemorrhage: measurements of cross-sectional areas of major blood vessels and volumes of aorta and spleen on MDCT and volumes of heart chambers on MRI. Am J Roentgenol 2006;187:209–15. 35. Jackowski C, Schweitzer W, Thali M, Yen K, Aghayev E, Sonnenschein M, et al., Virtopsy: postmortem imaging of the human heart in situ using MSCT and MRI. Forensic Sci Int 2005;149:11–23. 36. Jackowski C, Thali M, Aghayev E, Yen K, Sonnenschein M, Zwygart K, et al., Postmortem imaging of blood and its characteristics using MSCT and MRI. Int J Legal Med 2006;120:233–40. 37. Yen K, Lovblad KO, Scheurer E, Ozdoba C, Thali MJ, Aghayev E, et al., Post-mortem forensic neuroimaging: Correlation of MSCT and MRI findings with autopsy results. Forensic Sci Int 2007. 38. Poulsen K, Simonsen J, Computed tomography as routine in connection with medico-legal autopsies. Forensic Sci Int 2006. 39. Yen K, Vock P, Tiefenthaler B, Ranner G, Scheurer E, Thali MJ, et al., Virtopsy: forensic traumatology of the subcutaneous fatty tissue; multislice computed tomography (MSCT) and magnetic resonance imaging (MRI) as diagnostic tools. J Forensic Sci 2004;49:799–806. 40. Griffiths PD, Paley MN, Whitby EH, Post-mortem MRI as an adjunct to fetal or neonatal autopsy. Lancet 2005;365:1271–73. 41. Shiotani S, Yamazaki K, Kikuchi K, Nagata C, Morimoto T, Noguchi Y, et al., Postmortem magnetic resonance imaging (PMMRI) demonstration of reversible injury phase myocardium in a case of sudden death from acute coronary plaque change. Radiat Med 2005;23:563–65. 42. Widjaja E, Whitby EH, Cohen M, Paley MN, Griffiths PD, Post-mortem MRI of the foetal spine and spinal cord. Clin Radiol 2006;61:679–85. 43. Wolf RL, Zimmerman RA, Clancy R, Haselgrove JH, Quantitative apparent diffusion coefficient measurements in term neonates for early detection of hypoxic-ischemic brain injury: initial experience. Radiology 2001;218: 825–33. 44. Ward P, Counsell S, Allsop J, Cowan F, Shen Y, Edwards D, et al., Reduced fractional anisotropy on diffusion tensor magnetic resonance imaging after hypoxic-ischemic encephalopathy. Pediatrics 2006;117:e619–30. 45. Pallarito K, MRIs give their views to help hospitals mount defenses when they face malpractice charges. Mod Healthc 1992;22:43.
Incident-Specific Cases
46. Parizel PM, Demey HE, Veeckmans G, Verstreken F, Cras P, Jorens PG, et al., Early diagnosis of cerebral fat embolism syndrome by diffusion-weighted MRI (starfield pattern). Stroke 2001;32:2942–44. 47. Stanescu L, Talner LB, Mann FA, Diagnostic errors in polytrauma: a structured review of the recent literature. Emerg Radiol 2006;12:119–23. 48. Espinosa JA, Nolan TW, Reducing errors made by emergency physicians in interpreting radiographs: longitudinal study. BMJ 2000;320:737–40. 49. Milroy CM, Medical experts and the criminal courts. BMJ 2003;326:294–95.
© 2009 by Taylor & Francis Group, LLC
387
50. Stricker K, Orler R, Yen K, Takala J, Luginbuhl M, Severe hypercapnia due to pulmonary embolism of polymethylmethacrylate during vertebroplasty. Anesth Analg 2004; 98:1184–86, table of contents. 51. Yen K, Sonnenschein M, Thali MJ, Ozdoba C, Weis J, Zwygart K, et al., Postmortem multislice computed tomography and magnetic resonance imaging of odontoid fractures, atlantoaxial distractions and ascending medullary edema. Int J Legal Med 2005;119:129–36.
D4
Virtopsy as a Multi-Tool Approach
CONTENTS D4.1 Analysis, Visualization, and Reconstruction Methods ................................................................................................. 389 D4.1.1 Generation of 3D Models from Radiological Images ..................................................................................... 389 D4.1.2 Fusion of the Surface Data with the Internal Body Data................................................................................ 390 D4.1.3 Visualization of the Radiological CT and MRI Images to the Exterior Body (Cut with Virtual Knife, Virtual Cut) ............................................................................................................. 390 D4.1.4 Motion of the Virtual Body ............................................................................................................................ 391 D4.1.5 3D Reconstruction .......................................................................................................................................... 393 D4.1.6 Automatic Comparison of Injuries with the Assumed Injury-Inflicting Instrument ...................................... 393 D4.1.7 References ....................................................................................................................................................... 393 D4.2 Morphologic–Geometric Comparison of Patterned Injuries with the Assumed Injury-Causing Tool......................... 393 D4.2.1 Blunt Trauma .................................................................................................................................................. 393 D4.2.1.1 Case 1: Shoe .................................................................................................................................. 393 D4.2.1.2 Case 2: Baton................................................................................................................................. 397 D4.2.1.3 Case 3: Hammer ............................................................................................................................ 398 D4.2.2 Sharp Trauma.................................................................................................................................................. 399 D4.2.2.1 Knife .............................................................................................................................................. 399 D4.2.3 Muzzle Imprint ............................................................................................................................................... 400 D4.2.4 Bite Mark ........................................................................................................................................................ 403 D4.2.5 Strangulation ....................................................................................................................................................410 D4.2.6 References ....................................................................................................................................................... 411 D4.3 Forensic Reconstruction ................................................................................................................................................411 D4.3.1 Traffic Accidents ..............................................................................................................................................411 D4.3.1.1 Report Cases...................................................................................................................................412 D4.3.1.2 Collection of Evidence .................................................................................................................. 422 D4.3.2 Crime Scenarios .............................................................................................................................................. 429 D4.3.2.1 Report Case ................................................................................................................................... 429 D4.3.3 References ....................................................................................................................................................... 432 D4.4 Footprint Documentation .............................................................................................................................................. 432 D4.4.1 Introduction ..................................................................................................................................................... 432 D4.4.2 Method ............................................................................................................................................................ 434 D4.4.3 Results and Discussion.................................................................................................................................... 434 D4.4.4 References ....................................................................................................................................................... 435
D4.1 ANALYSIS, VISUALIZATION, AND RECONSTRUCTION METHODS Ursula Buck, Silvio Näther, Marcel Braun, and Michael J. Thali
D4.1.1 GENERATION OF 3D MODELS FROM RADIOLOGICAL IMAGES For the generation of 3D models of the osseous system and the skin surface of the deceased from plane multislice
computed tompography (MSCT) images, segmentation software is needed. At the Institute of Forensic Medicine in Bern, Switzerland, the software SliceViewer is used [1]. This program works with the threshold technique for the segmentation of medical 2D images into a 3D series. The respective pixels of the object are separated by an adjusted threshold value and are marked in a binary image. After the separation by threshold value, a new binary matrix for each image is created. The slice level allows for the computation of the 3D model from these created binary images (Figure D4.1.1). 389
© 2009 by Taylor & Francis Group, LLC
390
The Virtopsy Approach
A
B
FIGURE D4.1.1 (A) The 3D models of the skin and the bone structure of the deceased are generated from plane MSCT images by the software SliceViewer. (B) The CT images are imported into the software SliceViewer. With the adjustment of the threshold value, the requested data for the 3D modeling are extracted.
D4.1.2 FUSION OF THE SURFACE DATA WITH THE INTERNAL BODY DATA For the purpose of surface and radiological volume data set fusion, additional “radiological landmarks” [2] are stuck on human bodies. These markers are visible on the surface as
A
B
C
D
FIGURE D4.1.2 The 3D model of the external injury on the back and the osseous structure are fused to a real-data-based all-in-one 3D model of the abdomen region: (A) Skin model of the injury on the back in real color, generated from the data of the digital photogrammetry and 3D optical surface scanning. (B) Skin model of the back fused with the CT skin model of the front side. (C) Skin model of the back fused with the CT bone model. (D) Bone model generated from the CT images.
© 2009 by Taylor & Francis Group, LLC
well as on the radiological data set. They are used for the transformation of the surface model and the magnetic resonance imaging (MRI) data to the CT data by computer software. After the fusion, the correlation of the external injuries with the underlying injuries, which are visible in CT and MRI data, is possible. In cases where these markers are not used, the surface and radiological data are merged by using “anatomical landmarks,” such as prominent skin or bone structures. The merging or fusion process is actually made by specific 3D software programs. Figure D4.1.2 shows the result of a merged data set. The bones and the surface of the abdomen are generated from the CT images; the 3D model of the back of the body, with the injury in real color, is produced by the surface digitizing process.
D4.1.3 VISUALIZATION OF THE RADIOLOGICAL CT AND MRI IMAGES TO THE EXTERIOR BODY (CUT WITH VIRTUAL KNIFE, VIRTUAL CUT) For the examination and demonstration of internal injuries, it is necessary to visualize the CT and MR cross-section images (Figure D4.1.3). For the allocation of the relevant 2D CT and MR images to the 3D body model, including external
Virtopsy as a Multi-Tool Approach
A
B
FIGURE D4.1.3 An axial CT (A) and MR (B) image of the abdomen region. The MR and the CT image display the internal softtissue injury (red arrow). The radiological markers (yellow circle) are stuck onto the body for the fusion of the imaging data with the 3D surface scanning data.
injuries, the digital 3D body model is cut virtually, and the images are laid onto the cut surfaces by computer software (Figure D4.1.4 and Figure D4.1.5).
D4.1.4 MOTION OF THE VIRTUAL BODY To find the match between the injury and the injury-causing tool, the 3D model of the victim has to be brought into the probable positions at the time of the impact. Using the obtained real-data-based information of the osseous structure, especially the positions of the joints of the
391
A
B
FIGURE D4.1.4 Virtual cut through the 3D body model: Visualization of the internal autopsy findings combined with the external injuries of the corpse. The CT (A) and MR (B) image, laid onto the cut surface, display the internal findings on the left side.
corpse, the surface and skeleton model can be moved virtually in an anatomically correct manner in order to reconstruct the actual body position during the impact. For this reconstruction of the course of events, the 3D models of the corpse are moved with the software 3ds Max [3] using an adapted computer skeleton model (biped). The joints of this computer skeleton model are adapted to the individual bone structure of the CT model by scaling and rotating and then are covered with the skin model created from the CT and surface data (Figure D4.1.6A). Movements of the skeleton model effect the same movements in the CT models and the surface model (Figure D4.1.6B).
FIGURE D4.1.5 The all-in-one model of the corpse is cut virtually on different positions. This visualization shows the location of the internal and the external injuries, as well as the fractures of the ribs (red arrows).
© 2009 by Taylor & Francis Group, LLC
392
The Virtopsy Approach
A
B
FIGURE D4.1.6 Animation of the real-data-based 3D model: (A) The computer skeleton model (blue) is adapted to the individual body structure of the deceased. (B) The movement of the blue computer model effects the same movement of the 3D model of the deceased.
A
(mm) 5.0 4.0
B C
3.0 2.0 1.0 0.0 –1.0 –2.0 –3.0 –4.0 –5.0
FIGURE D4.1.7 Computer-aided geometric match analysis of overlapping footwear impressions in snow with the sole of a shoe of the suspected offender: (A) The differentiation of the individual impressions is possible (blue and red sole). (B) The deviation between the sole and the upper impression (from the red sole) is represented as a color-coded polygon mesh. The color scale represents the value of deviation. (C) A detailed view of the mesh shows that the green round structures match to the red sole (red arrows), and the blue round structures match to the overlapping position of the blue sole (blue arrows).
© 2009 by Taylor & Francis Group, LLC
Virtopsy as a Multi-Tool Approach
D4.1.5 3D RECONSTRUCTION The three-dimensional, geometric reconstruction is based on the true-color 3D models of the corpse and the presumed injury-causing tools. These true-to-scale 3D models are moved and rotated in a three-dimensional room by modern computer software to compare the injuries with the presumed injury-causing tool and to investigate the impact configuration. In cases of virtual reconstructions of incident scenes the 3D situation plans are integrated in the analysis and are imported into the virtual scenes. The aims of the virtual, real-data-based, and geometric reconstruction are the comparison of injuries on the skin and bones with assumed injury-causing instruments, the visualization of shot directions, and the determination of the impact positions and directions. The possibilities of the 3D geometric reconstruction are shown on real cases in Sections D4.2 and D4.3.
D4.1.6 AUTOMATIC COMPARISON OF INJURIES WITH THE ASSUMED INJURY-INFLICTING INSTRUMENT In the previously described computer-based manual comparison of accurate true-to-scale 3D models of injuries with assumed injury-inflicting instruments, the correlations are seen on the screen or in printed documentation. In some cases it is possible to verify the analysis with an automatic mathematical comparison. Therefore, the presumptive correlating meshes are transformed in the best fit position. The deviations between both meshes are automatically calculated by the software and are shown as a color-coded polygon mesh. The color scale represents the value of deviation. The automatic comparison is performed between the 3D models of injuries and injury-inflicting instruments, as well as between the 3D models of impressions in different materials and the presumed tool marks. Figure D4.1.7 shows an automatic comparison between a digitized impression in snow with the presumed sole of a shoe.
D4.1.7 REFERENCES Wirth J. and Jäger W., 1997. Voxel, Tomogramme und magnetische Bilder, Forschung Wissen Transfer, Heft 1, 14–15. Multi-Modality Markers for CT and /or MRI, IZI Medical Products, Baltimore, MD, USA. http://www.autodesk.com/3dsmax, U.S.A., 2007.
D4.2 MORPHOLOGIC–GEOMETRIC COMPARISON OF PATTERNED INJURIES WITH THE ASSUMED INJURY-CAUSING TOOL Ursula Buck, Silvio Näther, Marcel Braun, and Michael J. Thali The morphology of a patterned injury gives indications about the injury-inflicting instrument. It is rarely possible to
© 2009 by Taylor & Francis Group, LLC
393
identify the injury-inflicting instrument solely with the naked eye judging the morphology of the injury. In the majority of cases it is only possible to obtain information about the injury-causing tool after a detailed examination and analysis. A direct comparison of the presumed injury-inflicting instrument with the injury at the autopsy is almost impossible. Frequently at this time the presumed injury-inflicting instrument is unknown, and direct contact could mix important traces. The new process of 3D digitizing enables a one-toone comparison on the computer of the patterned injury with the injury-causing instrument. This method is noncontact and independent of time limitations. The virtual true-to-scale colored 3D models of the injury and the presumed instruments can be compared and contrasted by modern computer software. Consequently, evidence about the injury-inflicting instrument and how the instrument was used can be investigated. The analyses and results of real cases of injuries on the skin and bone of living and deceased persons caused by different injury-inflicting instruments are illustrated herein.
D4.2.1 BLUNT TRAUMA In everyday forensic practice the question often arises as to which injury-inflicting instrument caused the blunt trauma injuries to the victim. From this reconstructive point of view, forensic photogrammetry and surface scanning play an important role. With the computer-based correlation of the true-to-scale 3D models, which are produced in the Virtopsy Process, it is possible to link an injury-inflicting instrument to an injury [1–5]. This process and the results of the correlation are demonstrated on four real cases. D4.2.1.1 Case 1: Shoe Several persons had physically assaulted and kicked two men. The men had shoe-sole patterned injuries on the face, on the back, and on the neck (Figure D4.2.1). The injurycausing shoes had to be linked to the injuries out of seven shoes belonging to seven suspected offenders. All patterned skin injuries were documented using 3D photogrammetry at the hospital. The 3D model of the injury on the face is shown in Figure D4.2.2. The suspected shoes were digitized using the GOM surface-scanning system (Figure D4.2.3). The person who was kicked in the face was examined with a clinical computed tomography (CT) scan. By anatomical landmarks, the generated CT model from the clinical scan was fused with the 3D model of the part of the face with the external injury (Figure D4.2.4). Based on the true-to-scale 3D models, the patterned injuries of both men were compared with all suspected shoes and could be linked to the injury-causing shoes. Figure D4.2.5, Figure D4.2.6, and Figure D4.2.7 show the results of the match analysis of the patterned injuries with the suspected injury-causing shoes.
394
The Virtopsy Approach
FIGURE D4.2.1 Patterned injury on the forehead, on the back, and on the neck of a victim (Case 1).
FIGURE D4.2.2 The 3D model of the face. The ruler was placed on the head for the definition of the scale in the photogrammetric analysis (Case 1).
© 2009 by Taylor & Francis Group, LLC
Virtopsy as a Multi-Tool Approach
FIGURE D4.2.3 The resulting 3D models of the shoe soles of the suspected perpetrators (Case 1).
FIGURE D4.2.4 Fusion of the surface data of the exterior facial injury with the data of the clinical CT (Case 1).
FIGURE D4.2.5 3D match analysis of the patterned injury with the suspected injury-causing sole of a shoe (Case 1).
© 2009 by Taylor & Francis Group, LLC
395
396
The Virtopsy Approach
FIGURE D4.2.6 3D match analysis of the patterned injury and the suspected injury-causing sole of a shoe (Case 1).
FIGURE D4.2.7 3D match analysis of the patterned injury and the suspected injury-causing sole of a shoe (Case 1).
© 2009 by Taylor & Francis Group, LLC
Virtopsy as a Multi-Tool Approach
397
D4.2.1.2 Case 2: Baton
B
Another case of injuries on living persons was a victim who was physically attacked with a baton (Figure D4.2.8). The injury and the assumed injury-causing instrument were scanned using the ATOS III system. A photograph taken of the injury was projected onto the 3D skin surface. The 3D models of the patterned injury and of the baton displayed in detail the shape and color of the skin and instrument surfaces (Figure D4.2.9). A match analysis proved that the baton was considered as the injury-inflicting instrument (Figure D4.2.10 and Figure D4.2.11).
A
FIGURE D4.2.9 The 3D surface model of the injury (A) and the baton (B) generated by the surface scanning (Case 2). A
B
A
FIGURE D4.2.8 (A) The remarkable patterned injury left side of the back of a living person (Case 2). The black-white circular markers are stuck around the injury as reference targets for the automatic transformation of single measurements with the optical surface scanner. (B) The presumed injury-inflicting instrument was a flexible metallic baton.
FIGURE D4.2.10 The match analysis of the baton and the injury on the back. The shape of the patterned injury (A) matched with the shape of the baton (B) (Case 2).
FIGURE D4.2.11 The geometric match analysis shown from a 3D perspective (Case 2).
© 2009 by Taylor & Francis Group, LLC
B
398
The Virtopsy Approach
FIGURE D4.2.12 3D models of the three presumed injury-causing hammers (Case 3).
D4.2.1.3 Case 3: Hammer A man died due to fatal injuries to the head. He sustained 12 injuries on the head inflicted by an instrument with blunt force. The head injuries were documented using photogrammetry. Three hammers were found with the suspected offender, one of which was possibly the injury-inflicting instrument.
All three hammers were of approximately the same size and weight, but in detail they had individual forms and different signs of use (Figure D4.2.12). The computer-assisted comparison of the three hammers with the injuries resulted in the yellow hammer being matched to the patterned injuries (Figure D4.2.13 and Figure D4.2.14) and the blue and red hammers being excluded as injury-inflicting instruments.
FIGURE D4.2.13 Geometric comparison of the presumed hammers and the injuries on the head. The individual shape of the yellow hammer matched to the injuries. The shape of the blue and the red hammers did not correlate with the patterned injuries (Case 3).
© 2009 by Taylor & Francis Group, LLC
Virtopsy as a Multi-Tool Approach
399
?
FIGURE D4.2.15 Sharp force injuries to the skin, soft tissue and skeletal tissue (pelvis) (Case 4). Solely examining the skin injuries and the surface of the bone injury provided no clear clue to the type of knife used.
FIGURE D4.2.14 The comparison of the identified hammer with one of the patterned injuries showed the corresponding structures (Case 3).
In this case, the bone injuries were examined with a micro-CT system at the Institute of Medical Physics in Erlangen, Germany (Figure D4.2.16). The plane 2D images of the micro-CT volume data sets that optimally showed the wound profile of the knife were visualized in a computer-assisted design (CAD) program. For the comparison of the bone injuries with the knives, the thickness and form of the knives were exactly measured and drawn in the same software. The match analysis of the presumed knives with the wounds is shown in Figure D4.2.17, Figure D4.2.18, and Figure D4.2.19. The shape of the injury-inflicting knife was determined, and the match between bone injury and knife was proved.
D4.2.2 SHARP TRAUMA D4.2.2.1 Knife Case 4: Knife In a homicide case a man was stabbed with a knife. He had sharp force injuries to the skin, soft tissue, and skeletal tissue (pelvis). On the presumed knives there was no longer any DNA to link the injury to the injury-causing knife. Solely examining the skin injuries and the surface of the bone injury provided no clear clue to the type of knife used (Figure D4.2.15). When a knife penetrates bone, this sharp tool leaves an impression in the bone. The characteristics of shape and size indicate the type of injury-inflicting knife. To analyze patterned injuries of tool marks made in bone, micro-CT was used [5] and provided an opportunity to analyze patterned injuries in bone in a noninvasive manner. Using computer software a comparison between the bone injury and the presumed injury-causing instrument was possible.
© 2009 by Taylor & Francis Group, LLC
A
B
FIGURE D4.2.16 Digitalization of a bone part with an impression caused by a knife: (A) With a Micro-CT system at the Institute of Medical Physics in Erlangen, Germany. (B) 3D surface model of the scanned bone part, reconstructed out of the CT data (Case 4).
400
The Virtopsy Approach
FIGURE D4.2.17 Superimposition of a plane Micro-CT image, displaying the penetration injury of the bone part, and the CAD drawing of the presumable injury-causing knife (case 4).
FIGURE D4.2.19 A knife-induced injury to the bone with a rotational component (Case 4). Superimposition of the CT image and the CAD drawing of the knife.
D4.2.3 Muzzle Imprint In gunshot cases with muzzle imprints, it is important to know the position of the gun while firing the shot and to identify the injury-causing gun. To answer these questions an analysis of the wound is necessary. Case 5
FIGURE D4.2.18 A knife-induced injury to the bone with a rotational component (Case 4). Superimposition of the CT image and the CAD drawing of the knife.
The wound analysis is shown based on a muzzle imprint on the chest of a victim (Figure D4.2.20). 3D photogrammetry was used for the surface documentation in addition to the internal body morphology documentation using multislice CT (MSCT) and magnetic resonance imaging (MRI) techniques. The result of the 3D wound documentation was a 3D model of the external injury surface, resulting from the photogrammetric analysis (Figure D4.2.21), and a radiological
A
B
FIGURE D4.2.20 (A) Shotgun entrance wound on the chest of a victim, with a patterned injury around the entrance wound (muzzle imprint) (Case 5). (B) Photograph of the presumable injury-inflicting shotgun.
© 2009 by Taylor & Francis Group, LLC
Virtopsy as a Multi-Tool Approach
401
FIGURE D4.2.21 Photogrammetric analysis and visualization of the wound in 3D (Case 5).
data set of the internal volume of the body (Figure D4.2.22). Using the marked radiological landmarks, the internal and external data sets were merged together (Figure D4.2.23). Based on the geometric, true-to-scale data set of the muzzle imprint and the 3D model of the assumed injury-inflicting gun, the gun position while firing the shot was re-created (Figure D4.2.24).
FIGURE D4.2.22 CT cross-section through the gunshot injury in the 3D model of the body, generated from the CT images (Case 5). The multi-modality radiological markers (red lines) are used to give the local orientation between the cross-section and 3D model.
© 2009 by Taylor & Francis Group, LLC
FIGURE D4.2.23 Photogrammetric and radiological 3D data sets are merged into an all-in-one model including the body surface with the true-to-color muzzle imprint and the body inside (Case 5).
402
FIGURE D4.2.24 The reconstructed position of the gun during the shot phase (Case 5). In the 3D geometric analysis the muzzle imprint matches to the muzzle of the shotgun.
Case 6 This case showed a muzzle imprint on the right side of the head located above the ear (Figure D4.2.25) and an exit wound on the left side of the head. At the scene of the crime, two weapons were found, as well as 20 bullets that had been fired into a park bench next to the victim. The internal autopsy findings of the deceased were documented using the radiological techniques MSCT and MRI.
The Virtopsy Approach
FIGURE D4.2.25 3D surface digitizing of a muzzle imprint (Case 6): The scanner setup and the body, placed on the examination table in the autopsy room (top left-hand corner). Digital photograph of the muzzle imprint (big picture).
The external injuries on the head of the deceased, the presumed injury-inflicting gun and the projectile were digitized using the GOM ATOS II system. The gun was digitized when the slide was in the front and back positions, to capture the whole geometric form of the gun (Figure D4.2.26). The required 3D models of both positions were merged together by reference targets stuck on the gun. Then the gun slide was moved in the correct manner by the software. This was necessary to have all the positions of
FIGURE D4.2.26 3D surface scanning of the pistol (Case 6). 3D Digitizer ATOS II on a rotation table (left), the 3D model of the pistol with the slide pulled back and in front position (middle), a 3D model of the round.
© 2009 by Taylor & Francis Group, LLC
Virtopsy as a Multi-Tool Approach
403
match of the form of the muzzle of the gun with the injury lines was clearly visible (Figure D4.2.28). In addition to 3D models of the external injuries and the injury-inflicting gun, the 3D model of the head was virtually cut in axial and sagittal directions to show the bullet channel through the brain on the internal CT and MR images (Figure D4.2.29 and Figure D4.2.30). D4.2.4 Bite Mark
FIGURE D4.2.27 Based on the entrance and exit hole in the 3D model of the skull, the shot direction was reconstructed (Case 6).
the gun slide while a shot was being fired for the comparison with the muzzle imprint. Using the computer-assisted comparison of the 3D model of the muzzle imprint with the muzzle of the gun, the exact position of how the gun was held while firing the shot was ascertained. The shot direction was reconstructed with the entrance and exit hole of the bullet on the 3D model of the skull generated from the CT data (Figure D4.2.27). The
Bite marks are found, for example, on food, on malleable objects, or on skin. In this chapter the focus is on bite marks on the skin. Geometric bite-mark analysis for identification purposes is based on the individuality of a dentition, which is used to match a bite mark to a suspected perpetrator. This matching has to consider that biting is a dynamic process involving two moving systems. The mandible and the maxilla meet during the biting action. If the bite is applied to skin, the skin is moved too, depending upon the location of the part of the body and the tissue constitution at this site. Additional components of the bite pattern could be caused by the victim–biter interaction or movement.
Case 7 The bite-mark injury pattern on the back of a victim in a murder case showed the imprint of seven teeth of the mandible and five or six teeth of the maxilla (Figure D4.2.31(A)).
FIGURE D4.2.28 In the comparison of the 3D models the muzzle imprint could be matched with the pistol (Case 6). With the knowledge of the shot direction, the accurate position of the pistol on the head could be determined.
© 2009 by Taylor & Francis Group, LLC
404
The Virtopsy Approach
FIGURE D4.2.29 The position of the pistol during the shot was reconstructed using the data of 3D surface scanning and CT scanning (Case 6). The generated 3D models display the bone and skin defects; the plane CT images display the defects in the brain tissue.
FIGURE D4.2.30 The position of the pistol during the shot was reconstructed using the data of 3D surface scanning, and CT and MRI scanning (Case 6). The generated 3D models display the bone and skin defects, the plane CT images display the defects in the brain tissue. MR images have better tissue resolution. In the left-hand bottom corner, an MRI image combined with the 3D model of the head displays the shot channel.
© 2009 by Taylor & Francis Group, LLC
Virtopsy as a Multi-Tool Approach
405
A
C
B
FIGURE D4.2.31 3D documentation and analysis of a bite mark (Case 7). (A) 3D model of the bite mark generated 3D with photogrammetry data. (B) The dental casts of a suspected offender were digitized using the optical 3D surface scanner. The result is a 3D model of the casts of the maxillary and mandibulary dentition of the suspected offender. (C) The 3D model of the bite mark and the dentitions of the suspected offender were examined with respect to matching shapes, angles, and dimensions in the 3D CAD software (Case 7).
Several persons were suspected to be the perpetrator. Plaster casts of dentitions were manufactured and digitized using the 3D surface scanner from all suspected perpetrators (Figure D4.2.31(B)). The 3D models of the dentitions were matched against the bite mark on the skin in several steps to compare all teeth to the injury, to find out the different positions while biting, and to find out which dentitions did not match the bite mark (Figure D4.2.31(C)). An inside-out view of the 3D model of the bite mark, which is indicated by the red
boomerang arrow box, allowed a better visualization of the teeth entering the 3D model of the injury (Figure D4.2.32). This visualization means that the observer’s view is from inside the 3D model of the victim to the skin. This allows reverse-side analysis of the presumed matching areas of the dentitions with the skin lesions. Figure D4.2.33 shows how the teeth of a suspected perpetrator bit the skin with progressively increasing pressure. As is seen, there is a match at the beginning of the bite-mark
FIGURE D4.2.32 An inside-out view (symmetrical representation) of the 3D model of the bite mark, indicated by the red boomerang arrow box, allows for a better visualization of the development of the injury (Case 7).
FIGURE D4.2.33 Representation of the stepwise and dynamic process, in which the teeth penetrated the skin with progressively increasing pressure (Case 7). The match between the teeth and the patterned injury is marked with blue ellipses.
© 2009 by Taylor & Francis Group, LLC
406
The Virtopsy Approach
From the result of our photogrammetric investigation, it was strongly considered that the suspect whose cast is shown in Figure D4.2.31 was in fact the perpetrator. Subsequent DNA testing of the swab taken from the bite mark confirmed our photogrammetric conclusion. Case 8
FIGURE D4.2.34 The teeth of another suspected offender did not match with the bite mark, signaled with red ellipses (Case 7).
lacerations, where the front teeth first make contact with the skin. The tips of the premolars and molars follow exactly the scrape lesions as the bite progresses. Even in the lateral parts of the bite mark there are matching areas between the teeth and the injury. Also noticeable here are slight deviations in the lateral positions of the bite mark, which can be explained by the dynamic interaction between the teeth and the bitten tissue. An analysis of the dentition of another suspected perpetrator showed that these teeth did not match the bite mark (Figure D4.2.34).
In this homicide case a woman was stabbed with a knife. She died due to the fatal stab injuries. Furthermore, she was bitten on the right arm. The bite-mark injury was documented using TRITOP photogrammetry and the high-resolution surface scanner ATOS II (Figure D4.2.35). The 3D model of the bite mark was compared and contrasted with the 3D models of the plaster casts of the dentitions of two suspected perpetrators (Figure D4.2.36). The teeth of one suspected perpetrator (viewed in gray color) were not matched to the bite mark (Figure D4.2.37). The comparison of the wound with the dentition of the other suspected perpetrator resulted in the teeth of the mandible and maxilla being matched to the bite mark on the arm of the deceased (Figure D4.2.38, Figure D4.2.39, and Figure D4.2.40). The match of the front teeth of the mandible to the bite mark was visible on the true-color model of the skin, marked with orange, red, and blue lines, as well as on the gray-shaded 3D model of the wound. The 3D model showed the deepness of the impression on the skin, which correlated with the form of the teeth (Figure D4.2.41). Figure D4.2.42 shows the configuration of the perpetrator and the victim when her right arm was bitten.
FIGURE D4.2.35 3D documentation of a bite mark injury on the right forearm of the victim (Case 8).
© 2009 by Taylor & Francis Group, LLC
Virtopsy as a Multi-Tool Approach
407
FIGURE D4.2.36 The dentition casts of the suspected offenders were digitized and out of these data 3D models were generated (Case 8).
FIGURE D4.2.37 In the comparison of the surface model of the bite mark with the dentition models of the first suspected offender, a correlation could not be determined (Case 8).
FIGURE D4.2.38 In the comparison of the surface model of the bite mark with the dentition models of the second suspected offender, a correlation could be determined (Case 8).
© 2009 by Taylor & Francis Group, LLC
408
The Virtopsy Approach
FIGURE D4.2.39 In the comparison of the surface model of the bite mark with the dentition models of the second suspected offender, a correlation could be determined (Case 8).
FIGURE D4.2.40 In the comparison of the surface model of the bite mark with the dentition models of the second suspected offender, a correlation could be determined (Case 8).
© 2009 by Taylor & Francis Group, LLC
Virtopsy as a Multi-Tool Approach
409
FIGURE D4.2.41 The noncolored 3D model of the bite mark displays the impressions of the teeth (Case 8).
FIGURE D4.2.42 This graphic shows a possible position of the offender when biting the victim (Case 8).
FIGURE D4.2.43 A woman with claimed superficial patterned injuries on the neck (Case 9). The presumed injury-causing instrument was a necklace. The skin injury and the necklace were documented with the 3D surface scanner.
© 2009 by Taylor & Francis Group, LLC
410
The Virtopsy Approach
FIGURE D4.2.44 3D models of the skin injuries and the necklace (Case 9).
D4.2.5 Strangulation
Investigations of strangulation incidents of living or deceased persons are commonly performed in forensic medicine.
Besides the examination of the internal autopsy findings and collateral signs, the patterned injury on the skin can give clues to the strangulation tool. In this case a woman claimed that she was manually strangulated by an unknown person. She had superficial patterned
FIGURE D4.2.45 The match analysis of the necklace and the superficial skin injuries (Case 9).
FIGURE D4.2.46 The match analysis of the necklace and the superficial skin injuries (Case 9).
Case 9
© 2009 by Taylor & Francis Group, LLC
Virtopsy as a Multi-Tool Approach
injuries on the neck. The presumed injury-causing instrument was a necklace (Figure D4.2.43). The injuries on the neck of the victim and the necklace were digitized with the 3D surface scanner. A photograph was taken of the injuries on the neck and was projected onto the surface data. The result was a 3D model of the neck injuries and the necklace as shown in Figure D4.2.44. The match analysis concluded that the width of the links of the necklace could possibly fit the patterned injuries (Figure D4.2.45 and Figure D4.2.46), but the evidence was not conclusive enough because the necklace injuries were only superficial.
D4.2.6 REFERENCES Thali M.J., Braun M., Buck U., Aghayev E., Jackowski C., Vock P., Sonnenschein M., Dirnhofer D., VIRTOPSY—Scientific Documentation, Reconstruction and Animation in Forensic: Individual and Real 3D Data Based Geo-Metric Approach including Optical Body/Object Surface and Radio-logical CT/MRI Scanning. J Forensic Sci (2005) 50(2):428–24. Thali M.J., Braun M., Markwalder T.H., Brüschweiler W., Zollinger U., Malik NJ,Yen K., Dirnhofer R., Bite Mark Documentation and Analysis: The Forensic 3D/CAD Supported Photgrammetry Approach. Forensic Sci Int (2003) 135(2):115–21. Thali M.J., Braun M., Dirnhofer R., Optical 3D Surface Digitizing in Forensic Medicine. 3D Documentation of Skin and Bone Injuries. Forensic Sci Int (2003) 137(2–3):203–08. Thali M.J., Braun M., Wirth J., Vock P., Dirnhofer R., 3D Surface and Body Documentation in Forensic Medicine: 3D/CAD Photogrammetry Merged with 3D Radiological Scanning. J Forensic Sci (2003) 48(6):1356–65. Thali M.J., Taubenreuther U., Karolczak M., Braun M., Brueschweiler W., Kalender WA., Dirnhofer R., Forensic Microradiology: Micro-Computed Tomography (Micro-CT) and Analysis of Patterned Injuries Inside of Bone. J Forensic Sci (2003) 48(6):1336–42.
411
D4.3 FORENSIC RECONSTRUCTION Ursula Buck, Jörg Arnold, Silvio Näther, Marcel Braun, Herbert Bösch, and Michael J. Thali
D4.3.1 TRAFFIC ACCIDENTS For reconstruction of traffic accident events in cases of collisions between a motor vehicle and a pedestrian or cyclist, the correlation of injuries to the body with the injury-causing object and the accident mechanism are of great importance. Besides the photographic documentation and representation to scale of the accident scene and the involved vehicles as well as the investigation of traces, the results obtained by the forensic medical examination (external examination and autopsy) also give important information about the course of accident events [1]. The configuration of the vehicles on impact is an important aspect of the analysis. In addition to the exterior and interior body documentation of the deceased described in Chapter B1 are the accident vehicles and their damage documented by 3D optical surface scanning. The forensic analysis of traffic accidents includes the processing of the obtained data to 3D models 3D representation of the bone fractures and so the determination of the driving direction of the vehicle correlation of injuries with damage seen on the vehicles and geometric determination of the impact configuration as well as evaluation of further findings of the accident [2]. In the following cases, the methods and results of the described forensic medical 3D accident reconstruction are illustrated.
A
B
FIGURE D4.3.1.1 Due to the traces on the bicycle and the car, the position of both vehicles to each other could be reconstructed (Case 1): (A) The impact situation between car and bicycle seen from the right side. The damage on the bonnet of the car was caused by the handlebar of the bicycle. (B) Top view of the impact situation. The deformed bicycle frame corresponds to the shape of the bumper.
© 2009 by Taylor & Francis Group, LLC
412
The Virtopsy Approach
D4.3.1.1 Report Cases Report Case 1: Collision between a Car and a Bicycle
FIGURE D4.3.1.2 The colored 3D surface model of the accident victim displays the external impact injuries on both legs, the injuries on the left arm, and the head injuries (Case 1). The multimodality radiographic markers are used for the fusion of the external and internal data sets.
A
A boy was hit by a car while riding his bicycle from a harvested field onto a country road. He died several hours after the accident due to severe head trauma. Apart from cranial injuries, he also suffered several hematomas, bruises, and a fracture of the left tibia and fibula. The involved car displayed a dent and a scratch on the bonnet and damage to the right fender and to the right side of the front bumper. The bicycle showed a left-sided frame indention, a displaced saddle, and a contorted mudguard. The question of the collision configuration of the car, the bicycle, and the cyclist was addressed. In the classic procedure, the comparison of injuries and damage is only possible using 2D photographs, which only allows for an approximate correlation of significant patterned
B
C
WV
WW: 1241V
WW: 510WL: 337
FIGURE D4.3.1.3 The radiological MSCT/MRI cross-section images display the severe head trauma (Case 1): (A) In the axial CT image, a fracture of the left temporal bone reaching from the skull cap to the skull base is visible (red arrow). (B,C) A contusion of the scalp on the right temple and hemorrhages of the left temple are visible in the MRI sequences (yellow arrows). (a)
FIGURE D4.3.1.4 Real-data-based 3D model of the skin and bone structure (Case 1). The models are generated out of the CT data.
© 2009 by Taylor & Francis Group, LLC
(b)
FIGURE D4.3.1.5 Visualization of the impact injuries of the left lower leg (Case 1): (a) Clearly visible in the CT model: the wedge fracture (Messerer-fracture) of the tibia. (b) The colored surface model displays the impact injury (hematoma). For the fusion of surface data of the external injuries and CT data of the internal body, the surface model is transformed into the radiological coordinate system. Hereby multimodality markers (yellow arrows) serve as reference targets.
Virtopsy as a Multi-Tool Approach
413
FIGURE D4.3.1.6 Integrated three-dimensional skin and bone model (generated out of the CT data) in the assumed sessile position on the bicycle (Case 1).
FIGURE D4.3.1.7 Correlation of the vehicle parts and external impact injuries of the left lower leg (Case 1). The outward injuries were matched to the right side of the bumper of the car, and the inward injury was inflicted by the bicycle chain mounting (yellow arrow).
FIGURE D4.3.1.8 The displaced fracture fragments of the left tibia were virtually repositioned (Case 1). On the reconstructed model of the tibia, the convex fracture lines of the wedge fracture are visible (marked in the 3D model with red color). The yellow arrow tips in the evaluated car’s driving direction.
© 2009 by Taylor & Francis Group, LLC
414
The Virtopsy Approach
FIGURE D4.3.1.9 The geometric comparison of the wedge fracture of the left tibia with the bumper of the car. The comparison confirms that the car came from the left side and collided with the lower left leg while the cyclist was riding his bicycle (Case 1).
injuries. By computer-assisted 3D comparison of complex data sets, relevant findings regarding the course of accident events can be detected and visualized. On the basis of the 3D documented damage, the impact configuration of the car and the bicycle was re-created by computer. The scratch on the bonnet correlated to the han-
dlebar of the bicycle. The deformation of the bicycle frame matched the front of the car (Figure D4.3.1.1). After having investigated the positions of the involved vehicles to each other, the position of the victim was reconstructed. The colored 3D model of the exterior body of the victim showed patterned impact injuries on the legs, injuries on the left arm, and fatal head injuries (Figure D4.3.1.2). The
FIGURE D4.3.1.10 The 3D match analysis between the damage to the car and the injuries to the body indicates that the victim collided with the head on the bonnet (Case 1).
© 2009 by Taylor & Francis Group, LLC
Virtopsy as a Multi-Tool Approach
415
A
B
FIGURE D4.3.1.11 Reconstruction of the impact situation (Case 1): (A) After positioning the 3D models in the assumed position due to the traces, the bicycle wheels do not touch the ground. (B) The front of the car was lowered until the bicycle wheels had ground contact so that the reconstructing of the impact situation was completed. This reconstruction indicates that the driver of the car must have seen the cyclist and braked before the collision.
internal autopsy findings on the head were a fracture of the left temporal bone reaching from the skullcap to the skull base, a contusion of the scalp on the right temple, and hemorrhages of the left temple (Figure D3.4.1.3). The multislice computed tomography (MSCT) and magnetic resonance imaging (MRI) examinations, documenting
A
the whole body (Figure D4.3.1.4), were combined and visualized with the external findings (Figure D4.3.1.5). Based on the displaced saddle of the bicycle, it was assumed that the victim was seated. This assumption was validated by the injuries and the correlating injury-inflicting vehicle parts. For this comparison of the injuries with the injury-inflicting vehicles, the 3D model of the victim was placed onto the bicycle as seen in Figure D4.3.1.6. By performing this geometric comparison, the following contact points were allocated. The exact impact position was virtually reconstructed. The injuries of the outward lower left leg were attributed to the right side of the bumper of the car. The bicycle chain mounting (Figure D4.3.1.7) inflicted the inward injuries. The 3D model of the tibia was generated from plane MSCT images. For presentation purposes and for the reconstruction of the impact direction, each displaced fracture fragment was virtually repositioned (Figure D4.3.1.8). This reconstructed tibia displayed the classic wedge fracture with the convex fracture lines continuing to 0° to the base of the wedge. In such cases the base of the wedge points in the direction of the impact source, the tip pointing in the car’s driving direction. This medical finding confirmed that the car came from the left side and collided with the lower left leg of the deceased (Figure D4.3.1.9). This is what caused the damage to the right side of the front bumper. Afterward, the boy’s left hip and left shoulder collided against the right side of the bonnet, upon which he sustained the lethal head injuries that correlated with the dents on the bonnet (Figure D4.3.1.10). Subsequently, he slid from the car onto the road, thereby sustaining contusions of the right knee and the right shoulder. In this virtually reconstructed impact position, the bicycle did not have ground contact (Figure D4.3.1.11(A)). This was because the car had been scanned in an immobile state in contrast to the dynamics of the accident situation. Therefore,
B
5c
m
6,5 cm 3,
2c
m
C
FIGURE D4.3.1.12 3D models of the external injuries of the body of an accident victim (Case 2): (A) Impact injury on the outward right lower leg. (B) Impact injury on the right shoulder. (C) Injuries to the head caused by the secondary contact with the ground.
© 2009 by Taylor & Francis Group, LLC
416
The Virtopsy Approach
FIGURE D4.3.1.13 3D model of the skin and bone structure of the body generated out of the CT images. The yellow arrows signal the fracture of the right scapula, the right upper leg, and the foot. A
B
FIGURE D4.3.1.14 For the reconstruction of the impact situation in Case 2, the injury pattern was drawn onto the leg of a test person. (A) Injury lines on the leg of the deceased. (B) Injury lines drawn on the test person. A
B
the front of the car was virtually lowered until the wheels of the matched bicycle touched the ground. The newly created situation clearly indicated that the car was slowed down by braking at the time of the accident, which led to a compression of the car’s front shock absorber (Figure D4.3.1.11(B)). Report Case 2: Collision Between a Car and a Pedestrian In the next case a pedestrian was fatally injured in an accident with a car. Death occurred at the accident site due to blood aspiration in combination with a profound loss of blood, pneumothorax, and fat embolism. The body showed collision injuries on the right leg and shoulder. Externally visible injuries were seen on the left side of the head. Furthermore, he had an extensive trauma of the chest. It was initially unclear whether the man was knocked over, run over, or rolled over and in which position he was at the time of the collision. These questions were addressed by the accident reconstruction. In this case as well as the 3D documentation of the internal and external corpse, the front and underbody of the involved vehicle and the jacket worn by the man were digitized. In © 2009 by Taylor & Francis Group, LLC
FIGURE D4.3.1.15 (A) As seen in this figure the injury lines would not run parallel if the leg was in a straight position (Case 2). (B) If the leg is bent, the injury lines run parallel.
Virtopsy as a Multi-Tool Approach
417
FIGURE D4.3.1.16 The real-data-based reconstruction of the leg position during the impact (Case 2). (Left, middle) Surface data of the injury while the leg was straight. (Right) The injury lines on the virtually bent leg run parallel.
such cases, the external injuries are differentiated between impact and being rolled over or run over, as well as secondary injuries due to falling onto the road. A right-sided impact was assumed from the collision injuries of the skin in combination with excoriations and contusions of the subcutaneous fat tissue and muscles of the outward side of the lower right leg and the upper right arm (Figure D4.3.1.12). On the inner side of the left knee a large
excoriation with an underlying pocket-shaped shearing of the skin layers from the periosteum of the knee-cap (decollement) was seen [3,4]. There was a wedge fracture of the right femur and an amputation of the right leg as seen in the 3D model of the osseous structure (Figure D4.3.1.13). There was no fracture seen in the bone at the site of the external collision injury of the upper
FIGURE D4.3.1.17 Considering all injuries and traces, the impact situation could be reconstructed: the accident victim in a kneeling position (Case 2).
© 2009 by Taylor & Francis Group, LLC
418
The Virtopsy Approach
FIGURE D4.3.1.18 Damaged area of the left front corner of the car (Case 3).
right arm. However, a hyperintensity in T2-weighted fat-saturated MR sequence and a hypointensity in the T1-weighted MR sequence in the bone marrow of the right humerus were observed in the MRI images, indicating an edema or hemorrhage (contusion) there. According to Lysowski and Marek [5] this is a sign of an impact. This radiological finding in the right
humerus, in combination with the wedge fracture of the right femur and the soft-tissue injuries, confirmed the assumption that the collision occurred to the right side of the pedestrian. That he was knocked over in an upright or bent position before the car subsequently ran or rolled over him was proved by the following findings: a wedge fracture of the femur, a
FIGURE D4.3.1.19 The motorcycle with damages on the right side is prepared with reference targets and scale bars for the photogrammetrical documentation (Case 3).
© 2009 by Taylor & Francis Group, LLC
Virtopsy as a Multi-Tool Approach
419
A
C B
D
FIGURE D4.3.1.20 3D model of the left front corner of the car. The red arrows point to different damage areas. The black lines signal the different heights above the ground (Case 3).
E
F
FIGURE D4.3.1.21 3D model of the damaged area of the motorcycle (Case 3).
© 2009 by Taylor & Francis Group, LLC
traumatic amputation of the right leg, and an abrasion of the right sole of the shoe [6]. The 3D model of the vehicle digitized by the surface scanner showed that the front part of the car had only parallel and rectangular structures. When the 3D model of the injuries of the right leg was examined closely, it became evident that if the leg had been in a straight position, the injury lines would not be parallel (Figure D4.3.1.14(A)). Therefore, the injuries could not have been sustained if the leg was in a straight position. If the leg had been bent, the injury lines would run parallel to each other and the vertical abrasion would form a right angle to the horizontal lines. Thus, the flexion degree of the leg at the time of the collision could be determined. This situation was reconstructed on a test person. The injury lines were drawn onto the leg (Figure D4.3.1.14(B) and Figure D4.3.1.15). Figure D4.3.1.16 shows the reconstruction of the 3D model of the real-data-based right leg. To bring the 3D data set of the body in the assumed impact position, the computer model (biped) was adapted to the skeleton model of the body. The precise impact position
420
The Virtopsy Approach
B
A
K J l H F G
FIGURE D4.3.1.22 High-resolution 3D model of the motorcycle (Case 3). Damaged prominent part of the motorcycle in top view. In this front view of the 3D model of the damaged area the height of the different parts (red arrows) is represented by the black lines.
ca .3 ° 5
than the height of the car bonnet, he was rolled over and not thrown onto the car (Figure D4.3.1.17) [7]. Report Case 3: Collision between a Car and a Two-Wheel Vehicle
FIGURE D4.3.1.23 It could be shown clearly that the motor bike collided with the car in an almost upright position.
was examined with respect to the vehicle front, the injury pattern of the right leg, and of the upper right arm (Figure D4.3.1.17). The décollement of the left knee indicated a tangential, shifting force during the right-side impact while the man was kneeling on his left knee. The fracture of the right femur was then considered. Different positions were adopted by the model until the injury pattern matched the vehicle structures. The décollement, which was seen on the back, and the matching tire imprint on the jacket worn by the man proved that the victim was rolled over. Ultimately, it could be proven that the man was struck by the car from the right side while kneeling on the tarmac and leaning on his right arm. As his center of gravity was lower
© 2009 by Taylor & Francis Group, LLC
In the next case the main purpose of a detailed 3D examination of the involved vehicles was to check two contrary scenarios given by two involved drivers. A motorcyclist was seriously injured in a car accident. Paraplegia occurred due to multiple traumata: After the initial collision between the motorcycle and the car, the motorcyclist was thrown in front of a small truck and was overrun by the truck. The driver of the car fled the scene in his car. The motorcyclist could not be interviewed for several weeks due to injuries. The next day the driver of the car showed up at a police station and mentioned the serious accident in the following way. He had seen a collision between the small truck and the motorcycle where the motorcyclist was overrun by a truck. The motorcycle was thrown away and collided finally with his car and damaged it. The next morning he found the damages on his car to be worse than expected and wanted them to be paid for by the insurance of the two other involved vehicles. Because of the severe differences between the version of the truck driver and the version of the car driver, the public prosecutor wanted an accurate reconstruction of the collision configurations between the involved vehicles. The inspection of the car showed a damaged area at the left front corner of the vehicle where at least three different contacts must have happened (Figure D4.3.1.18). The height of the uppermost damage areas on the two vehicles was too high to fit the scenario mentioned by the car driver. The whole front of the car was photographed and scanned with the 3D surface scanner. All physical evidence (e.g.,
Virtopsy as a Multi-Tool Approach
microtraces like fibers, paint chips, plastic smears, and biological material that eventually were sticking on the car’s surface) was collected with the taping method. The same procedures were applied to the damaged motorcycle (Figure D4.3.1.19). A special interest was given to the different outermost parts of the motorcycle such as the handlebar, the protective covers on the exhaust pipe, footrests, and the front fork. The superposition of the two 3D models of the car and the motorcycle showed clearly that the accident had happened in a completely different way from what the car driver had told the police. The primary collision occurred between the left front corner of the car that left a road without right of way and blocked the arriving motorcycle’s way. The motorcycle collided with its right side against the left front corner of the car. The different damage areas showed corresponding morphological properties, and their height above the ground corresponded with very small differences (Figure D4.3.1.20, Figure D4.3.1.21, and Figure D4.3.1.22). The collision configuration could be reconstructed with respect to the relative positions and the angles between the two vehicles. In addition, it could be shown clearly that the motorcycle collided in an almost upright position with the car (Figure D4.3.1.23). The motorbike was redirected
421
toward its left side and as a result of the first impact was thrown in the direction of the small truck and hit the road surface. Finally, the driver of the motor bike was overrun by the approaching truck. Report Case 4: Accident-Site Documentation in a Traffic Accident at High Speed and Complex Multivehicle Collision Traffic accidents at high speed are very often accidents with more than two involved vehicles and multiple collisions. Additional difficulties arise in accident reconstruction due to the severe damages and massive deformations that will be found on the vehicles. If vehicles are involved in more than one collision, the estimation of the absorbed energies in the collisions becomes very difficult, as crash tests normally are performed with undamaged cars. As soon as a car is already deformed when it enters into a second or third collision, it will react differently from an undamaged car. Many questions arise and must be answered—often with no or very little knowledge from eyewitnesses. Usually such accidents occur without or with almost no warning to the involved persons. Thus, the judges and the assurances
FIGURE D4.3.1.24 Photographic documentation of the accident site, the damaged vehicles, and the position of the victims (yellow ellipses) (Case 4).
© 2009 by Taylor & Francis Group, LLC
422
The Virtopsy Approach
FIGURE D4.3.1.25 Collected microtraces from the different vehicle parts (Case 4).
depend essentially on the results of accident reconstructions to examine such an event appropriately: r What happened at the accident site and in which temporal order? r What were the original directions of travel, and what were the initial velocities of the different vehicles? r What were the different collision configurations? r Could the drivers have prevented any of the collisions? r Could the drivers have reduced the resulting damages or injuries if they had behaved in a different way? It was a Saturday night on a Swiss freeway. As a result of a traffic accident with multiple collisions in which three vehicles were involved, the local police found three heavily damaged cars and two killed persons in their final positions after the accident on the freeway (Figure D4.3.1.24). The development of the accident was not clear.
© 2009 by Taylor & Francis Group, LLC
D4.3.1.2 Collection of Evidence During the night after the accident and during the following morning, no photogrammetry equipment was available. The freeway police and the local forensic service made a detailed photo documentation and secured the three damaged vehicles and the clothing of the two victims. At the accident site, many relevant distances were measured. The forensic service collected microtraces (e.g., paint, plastic traces, fibers, glass fragments) from all visible collision damages on the side rails of the freeway. The Scientific Forensic Service was engaged to complete the collection of evidence on the vehicles and on the clothing of the victim and to perform a thorough reconstruction of this traffic accident. The essential questions to be answered were “Which vehicle collided with which victim?” and “How fast were the vehicles traveling before the collisions?” Microtraces from all vehicles, the illumination units and the bulbs from the different vehicle parts, and material samples were collected for comparison purposes (Figure D4.3.1.25). On the vehicles we documented blood spatter patterns and collected DNA samples. The most important data for the public prosecutor was a scale plan of the accident situation. On a Sunday several weeks after the accident had happened, we could document the whole accident area on the
Virtopsy as a Multi-Tool Approach
423
FIGURE D4.3.1.26 Subsequent evaluation of noncalibrated standard pictures. With this method tire tracks, collision sites, and positions of victims could be integrated in the situation plan (Case 4).
FIGURE D4.3.1.27 Situation plan with the complete set of found traces (Case 4).
© 2009 by Taylor & Francis Group, LLC
424
The Virtopsy Approach
FIGURE D4.3.1.28 3D point cloud—3D triangulation: The front of the Audi is shown as rendered 3D point cloud and the rear end as 3D triangle grid after the triangulation (Case 4).
FIGURE D4.3.1.29 3D model of the damaged BMW (Case 4).
FIGURE D4.3.1.30 3D model of the damaged front of the Lancia (Case 4).
© 2009 by Taylor & Francis Group, LLC
Virtopsy as a Multi-Tool Approach
425
FIGURE D4.3.1.31 Collision configuration of the first impact: The reconstructed collision configuration in the moment of the deepest intrusion during the first collision. We found matching damages and detail traces on the corner of the motor hood of the BMW and on the motor hood of the Audi and the impression of a wheel nut of the BMW in the front beam of the Audi.
closed freeway together with our photogrammetry specialists with the RolleiMetric system. With these photos an accurate 3D model of the whole accident area could be calculated. During subsequent evaluation of noncalibrated standard pictures, using the fact that many objects in the whole accident area are unchanged before, during, and after the accident allows the noncalibrated standard pictures to be fit. The noncalibrated standard pictures are “moved” through
the 3D model of the accident site until the unchanged reference points match. As soon as the match is established, it is possible to evaluate and measure the additional information that is visible on the pictures of the accident site (e.g., tire marks, final positions). Using this method we could integrate a large number of tire marks, collision sites on the side rails, fields of debris, and final positions of victims and vehicles on scale onto the situation plan (Figure D4.3.1.26).
FIGURE D4.3.1.32 Collision development: Additional matching damages on the front doors of the two vehicles could be evaluated to reconstruct the further collision development. It was even possible to reconstruct the accurate position of the driver’s door of the BMW during the collision of the Lancia according to matching damages.
© 2009 by Taylor & Francis Group, LLC
426
The Virtopsy Approach
FIGURE D4.3.1.33 3D scanning of undamaged cars for comparison purposes.
FIGURE D4.3.1.34 Comparison of the 3D models of the accident car with the undamaged car to measure the intrusion depth.
© 2009 by Taylor & Francis Group, LLC
Virtopsy as a Multi-Tool Approach
427
1 B+C A
0m 10 m
80 m
20 m
90 m
30 m
100 m
40 m
50 m
120 m
110 m
60 m
130 m
70 m
140 m
2
FIGURE D4.3.1.35 Detailed reconstruction of the accident: It started with loss of control of the BMW, whose driver was speeding and probably made an abrupt maneuver.
A
B 180 m
190 m
200 m
FIGURE D4.3.1.36 Detailed reconstruction of the accident (A) The collision position of the BMW and the Audi. (B) When the police arrived at the accident site, these final positions could be found.
© 2009 by Taylor & Francis Group, LLC
428
The Virtopsy Approach
FIGURE D4.3.1.37 3D model of the accident scene including the tire tracks, the positions of the cars during the accident, and the final positions of the victims.
The result was a detailed situation plan covering the whole accident area with a length of more than 300 m. A series of collision sites with the side rails and the final positions of the vehicles and the victims allowed a first rough reconstruction of the accident development (Figure D4.3.1.27). The missing pieces of information were the accurate collision sites and the collision configurations between the vehicles or the victims.
For further analysis, the three accident vehicles were completely scanned in 3D. The results of the digitizing process with the TRITOP system and ATOS II scanner are shown in Figure D4.3.1.28, Figure D4.3.1.29, and Figure D4.3.1.30. After regrouping the point cloud into reasonable objects and moving the 3D models in the virtual space, matching damages could be fitted to find the spatial orientation of the vehicles in the moment of the deepest intrusion during
FIGURE D4.3.2.1 3D model of the crime scene of a gunshot case. The direction of the trajectory of the bullet was determined by tachymetry.
© 2009 by Taylor & Francis Group, LLC
Virtopsy as a Multi-Tool Approach
429
FIGURE D4.3.2.2 The corpse was documented in 3D using the CT scanner, system and skin surface. Based on these models, a computer animation model was adapted to the individual physique of the deceased.
the collision (Figure D4.3.1.31 and Figure D4.3.1.32). In addition to the accurate positions and orientations of the vehicles during the collision, it was important for the accident reconstruction to know the intrusion depths during the collisions. For the measurement of the intrusion depths in detail and in any chosen direction, undamaged vehicles of the same types were digitized (Figure D4.3.1.33 and Figure D4.3.1.34). The deformations can be compared with deformations from crash tests. This enables us to estimate the amount of energy that was transformed from movement into deformations. After the evaluation of all the collected evidence, we could reconstruct the whole development of the accident (Figure D4.3.1.35 and Figure D4.3.1.36). After the loss of control of the BMW and several collisions with the side rails, the BMW came to a first stop in the left lane. The two passengers left the BMW and started to walk back on the freeway. When the Audi arrived at the accident site, the Audi collided first with the two pedestrians and immediately after with the BMW. Shortly after that one of the pedestrians was hit by the arriving Lancia and dragged along, whereas the BMW hit on his way the other pedestrian, who also was dragged along. The Lancia collided finally with the driver’s side of the BMW. All photogrammetrically evaluated detail situations, including the traces, final positions, and damage of the cars, can be combined and fitted together to visualize the development of the accident. Using all the reconstructed intermediate positions, the accident simulation could be done with high accuracy. The calculations based on the reconstructed accident development resulted in the initial speeds of the three cars. The speed tolerances could be kept very narrow, because the evidence-based reconstruction was already very precise. As a final result we could prove (based on the collected evidence from the accident site, the vehicles, and the victims) that the two passengers of the BMW had left the vehicle after
© 2009 by Taylor & Francis Group, LLC
the first accident. Then, both “pedestrians” were hit at a high speed by the arriving Audi (Figure D4.3.1.37).
D4.3.2 CRIME SCENARIOS D4.3.2.1 Report Case Report Case 5: Digital Bullet Trajectory Reconstruction in Gunshot Cases In gunshot cases the reconstruction of the bullet trajectory has two components: (1) the bullet channel in the body of the deceased (see Chapter D4.2), and (2) the direction of the bullet trajectory at the scene of the crime if the bullet went through
FIGURE D4.3.2.3 The computer model of the corpse positioned on the bed in the 3D room. This was the end position in which deceased was found.
430
The Virtopsy Approach
FIGURE D4.3.2.4 3D model of the crime scene in a view of a different perspective.
objects. The investigation of both components in one analysis gives clues to the configuration of the offender and victim and where the crime occurred. In addition, visualizations of how the act of the crime was seen from the site and the angle of a witness
or a person’s view are possible. For this approach, cooperation between police and forensic medicine is of great importance. A reconstruction of a homicide case by gunshot analyzed by the police and forensic medicine experts is as follows:
FIGURE D4.3.2.5 The presumable offender–victim configuration, when the pistol was fired, was reconstructed considering the traces, the physiques of the victim and offender, the bullet holes in the mattress and head of the bed, and as the injuries of the victim.
© 2009 by Taylor & Francis Group, LLC
Virtopsy as a Multi-Tool Approach
A father was killed by his son in the parents’ bedroom from a gunshot to the head. The shot was heard by the mother from another room in the house. She found her husband lying on the bed and her son standing in the room with a gun in his right hand. The ambulance team could only diagnose the death. The son gave evidence that in the course of showing the functions and components of the gun to the father a projectile was discharged because of a faulty function of the gun. A detailed geometric analysis was carried out to discover whether it was a tragic, accidentally-fired shot or a criminal assault. The autopsy finding was an entrance wound to the left side of the nose and an exit wound on the right side of the nape of the neck. The bullet went through the soft tissue on the right side of the cervical spine. The man died due to loss of blood caused by the gunshot injury to the head. At the crime scene the body of the deceased was lying on the bed, and there were a bloodstained quilt and mattress and a bullet channel through the mattress as well as a ricochet mark caused by the bullet in the wooden headboard. The bullet and the gun were sited on the top of a cat’s climbing frame next to the door.
431
The crime scene was documented by the police experts in 3D using photogrammetry and tachymetry. The direction of the trajectory of the bullet through the mattress was measured by means of a rod with a mounted reflector that has a known length and the same diameter as the bullet, which was 9 mm. This reflector rod was put through the channel of the mattress with contact on the ricochet mark in the headboard. The angles and distances were measured using the tachymeter. A true-to-scale replicated crime scene 3D model was drawn by computer-assisted design (CAD) software using the results of the photogrammetry analysis and the measurements required by the tachymetry. This created 3D model of the crime scene included all important information for the investigation (Figure D4.3.2.1). As is usual in the Virtopsy Project, the corpse was documented using the radiological techniques multislice computed tomography (MSCT) and magnetic resonance imaging (MRI). The entrance and exit wounds were photographed. The 3D model of the deceased generated from the CT images displayed the entrance and exit wounds on the head. The bullet channel through the head of the deceased was drawn as a result of this CT model. The 3D model of the corpse was moved with computer software to bring the corpse into the position it was found in and the probable positions at the time of the firing of the shot. Thereby a computer skeleton model was adapted to the
FIGURE D4.3.2.6 3D model of the presumable offender–victim configuration in a view of a different perspective.
© 2009 by Taylor & Francis Group, LLC
432
The Virtopsy Approach
individual bone structure of the CT model of the corpse (Figure D4.3.2.2). The computer animation model of the offender was adapted by using his body measurements. The next step was the fusion of the 3D scene with the models of the involved persons to reconstruct what happened. Thereby all modeled data were combined in one file in the animation software 3ds Max. [8] The end situation of the scene of crime was reconstructed based on photographs that were taken by the police officers at the scene. These photographs were integrated into the photogrammetric analysis. After computing the camera positions and the calibration data of the camera used, the photographs were projected in the 3D room by the animation software. Then the computer model of the corpse was moved into the precise posture and position of the deceased lying on the bed based on the projected photographs (Figure D4.3.2.3 and Figure D4.3.2.4). The shooting was virtually reconstructed based on the directions of the bullet trajectory in the bedroom and through the head of the victim. The computer model of the corpse was moved to bring it into the probable position when the bullet channel through his head matched the direction of the bullet trajectory in the room. For the examination of the probable positions of the shot, the computer model of the offender was moved in the 3D scene (Figure D4.3.2.5 and Figure D4.3.2.6). In this reconstructed offender–victim configuration, the measured distance between gun muzzle and entrance wound of the victim was equal to the distance in the smoke residue analysis results.
D4.3.3 REFERENCES 1. D. Metter, Rechtsmedizinische Unfallrekonstruktion von tödlichen Fussgänger-PKW-Unfällen. Z Rechtsmedizin (1983) 91:21–23.
2. U. Buck, S. Naether, M. Braun, S. Bolliger, H. Friederich, C. Jackowski, et al., Application of 3D Documentation and Geometric Reconstruction Methods in Traffic Accident Analysis: With High Resolution Surface Scanning, Radiological MSCT/MRI Scanning and Real Data Based Animation. Forensic Sci Int (2007) 170:20–28. 3. D. Metter, Das Decollement als Anfahrverletzung. Z Rechtsmedizin (1980) 85:211–19. 4. O. Prokop and G. Radam, Rekonstruktion von Verkehrsunfällen aus gerichtsärztlicher Sicht. Handbuch der Verkehrsmedizin, 1968, S. 952., W. Weimann, Radspuren an Überfahrenen und ihre kriminalistische Bedeutung. Arch Kriminol (1927) 80:1–6. 5. Z. Lisowski and Z. Marek, Hyperpigmentation in the Long Bones of the Lower Limbs as a Basis for Vehicle Identification and Traffic Accident Reconstruction. Forensic Sci Int (1982) 20:251–55. 6. B. Karger, K. Teige, M. Fuchs, AND B. Brinkmann, Was the Pedestrian Hit in an Erect Position before Being Run Over? Forensic Sci Int (2001) 119:217–20. 7. U. Löhle, Zur Unfallrekonstruktion bei verschiedenen Unfalltypen. AJP/PJA (1995) 4. 8. http://www.autodesk.com/3dmax, USA, 2007.
D4.4 FOOTPRINT DOCUMENTATION Ursula Buck, Silvio Näther, and Michael J. Thali
D4.4.1 INTRODUCTION The 3D documentation of footwear and tire impressions in snow offers an opportunity to capture additional fine details for identification than present 2D photographs. For this approach of acquiring the depth of the impressions in snow, different casting methods are used. Traditional methods of casting use a kind of gypsum or sulfur in addition to Snow Print Wax.
FIGURE D4.4.1 Photograph of the shoe impression in snow with a ruler. The impression was highlighted with Snow Print Wax to increase the contrast.
© 2009 by Taylor & Francis Group, LLC
Virtopsy as a Multi-Tool Approach
A
B
433
C
FIGURE D4.4.2 Results of the 3D optical scanning: Accurate, highresolution 3D models (Experiment 2) of (A) the shoe, (B) the ceramic plaster casting, and (C) the optically scanned impression in snow.
A
B
Casting of footwear impressions in snow has always been a difficult assignment [1]. Casting with gypsum is risky, because important details for characteristic identification may be destroyed through the weight of the gypsum and by exothermic reactions occurring while the gypsum hardens. This risk can be reduced by using Snow Print Wax [2], gray primer [3], or dry powder layers such as Quickrete [4]. For casting with sulfur, the following difficulties exist. If the temperature of the sulfur is too high when it flows into the impression, important details of the snow impression are destroyed. Furthermore, in loose, powdery, or frozen snow the sulfur has a tendency to flow through the impression and collect under the surface [5]. Another difficulty in casting is that low temperatures complicate the process of casting and have a negative influence on the results [6,7]. Finally, after casting the impression is destroyed. For a noncontact approach of 3D documentation of impressions in snow at the scene of the crime, 3D surface scanning is a suitable and practical method. The surface scanning method for the 3D documentation of impressions in snow was evaluated and validated. The conventional methods of C
FIGURE D4.4.3 Analysis of wear and identifying characteristics with a side-by-side comparison: (A) Measuring dimensions in the 3D model of the sole of the shoe for comparison with the impression. The yellow arrow signals an accidental detail. (B) The profile depth in the 3D model of the gypsum cast is 2 mm. The surface of the cast is porous. Many details are not visible. (C) The profile depth in the 3D model of the impression is identical to the one in the original shoe profile (red arrow). The yellow arrow marks a potentially identifying characteristic.
FIGURE D4.4.4 Comparison of the impression in snow with the shoe.
© 2009 by Taylor & Francis Group, LLC
434
The Virtopsy Approach
FIGURE D4.4.5 Automatic comparison of the 3D models of the impression and the scanned sole by the ATOS software. The deviation between shoe and impression is represented by a colored polygon mesh. The value of deviation is represented by the color scale.
casting footwear impressions in snow with gypsum and sulfur were performed and compared with the new method. This was done in three experiments in real outdoor situations in different weather conditions and in different types of snow [8].
D4.4.2 METHOD For photographing and scanning it was sufficient to highlight the impressions with a gray spray to increase the contrast [9]. The use of Snow Print Wax is necessary when casting with gypsum in order to prevent the negative influence of an exothermic reaction (Figure D4.4.1) [10]. The 3D documentation of the impressions in snow at the scene of the crime and the presumed crime-inflicting shoes and car tires was performed using the GOM ATOS II system (GOM, Braunschweig, Germany) (Figure D4.4.2 and Figure D4.4.3). These produced accurate, true-to-scale 3D models of the impressions in snow, and the presumed crime-inflicting shoes or tires were compared with each other using the ATOS software (Figure D4.4.4). In addition to the side-by-side comparison method, to find identifying characteristics, an automatic superimposition comparison of impression and sole or tire is possible. In this way the presumed correlation meshes of the snow impression and the sole of a shoe were transformed in the best fit position.
© 2009 by Taylor & Francis Group, LLC
The deviations of the meshes were automatically computed by ATOS software and were shown as a color-coded polygon mesh. The color scale represented the value of deviation. A one-to-one comparison of the whole snow impression with the sole did not correlate, because the generation of the impression (the footstep) was a dynamic process. When walking over a soft, yielding surface, the sole of a shoe is impressed in a mound [1]. Based on the mechanics of a common scenario in three-dimensional impressions, the various positions of a shoe while the rolling foot is touching the ground could be determined with the automatic comparison utility of the ATOS software. Figure D4.5.5 shows the result of the geometric comparison between an impression and the matching sole of a shoe.
D4.4.3 RESULTS AND DISCUSSION The 3D models of the scanned impressions displayed the snow surface in high resolution, including all fine details. This method delivered more detailed results of a higher accuracy than the conventional casting techniques. Best results were achieved even in different kinds of snow and under different meteorological conditions. The optical surface scanning method is nondestructive, and the process can be repeated if errors occur during the scanning, whereas the classical casting techniques destroy the
Virtopsy as a Multi-Tool Approach
impression and mistakes cannot be rectified. Furthermore, the computer-assisted comparison between the impressions in question and the sole of a shoe of the suspected perpetrator offers high efficiency and new possibilities. In addition to the side-by-side comparison, the automatic comparison of the 3D models and the computation of deviations and accuracy of the data simplify the examination and deliver objective and secure results. The results can be presented as a demonstrative expertise. The method is also suitable for impressions in soil, sand, or other materials.
435
5.
6.
7.
8.
D4.4.4 REFERENCES 1. William J. Bodziak, Footwear Impression Evidence. CRC Press, Boca Raton, FL, 1995, pp. 63–101. 2. SM Ojena, A New Improved Technique for Casting Impressions in Snow. Journal of Forensic Science 29(1):322–25, 1984. 3. JR Wolfe and CW Beheim, Dental Snow Casting of Snow Impressions. In: Advances in Forensic Science, vol. 4, Berlin, 1995. 4. TW Adair, S Hisey, and R Tewes, Casting Snow Prints with “Quikrete” Fast Setting Concrete: An Alternative to Aerosol
© 2009 by Taylor & Francis Group, LLC
9.
10.
Wax Products. Information Bulletin for Shoeprint/Toolmark Examiners 11(1), April 2005. K Carlsson and AC Maehly, Nouvelles méthodes pour relever les empreintes de chaussures et de pneus sur différentes surfaces. International Criminal Police Review 199:158–67, 1976. L Nause and JC Forsythe-Erman, Casting Footwear Impressions in Snow: Snowprint-Wax vs. Prill Sulphur, RCMP Gazette 54(12), 1–7, 1992. JS Brennan, Dental Stone for Casting Depressed Shoemarks and Tyremarks. Journal of Forensic Science Society 23, 275–86, 1983. U Buck, N Albertini, S Naether, and MJ Thali, 3D Documentation of Footwear Impressions and Tyre Tracks in Snow with High Resolution Optical Surface Scanning. Forensic Science International 171, 157–64, 2006. E Hueske, Photographing and Casting Footwear/Tiretrack Impressions in Snow. Journal of Forensic Identification 41(2), 92–95, 1991. L Hammer and J Wolfe, Shoe and Tire Impressions in Snow: Photography and Casting. Journal of Forensic Identification 53(6), 647–55, 2003.
D5
Biopsy Steffen Ross, Emin Aghayev, and Michael J. Thali
CONTENTS D5.1 Introduction .................................................................................................................................................................. 437 D5.2 Historic Development .................................................................................................................................................... 437 D5.2.1 Non-Image-Guided Biopsy and Needle Autopsy ............................................................................................. 437 D5.3 Targeting Procedures for Correct Needle Placement .................................................................................................... 439 D5.3.1 Ultrasound-Guided Biopsy ............................................................................................................................... 439 D5.3.2 Magnetic Resonance Imaging .......................................................................................................................... 439 D5.3.3 CT-Guided Biopsy ............................................................................................................................................ 439 D5.3.3.1 Classic Step-and-Shot Technique ................................................................................................... 439 D5.3.3.2 CT Fluoroscopy ............................................................................................................................... 439 D5.3.3.3 Navigated Biopsy with Robotic Assistance ..................................................................................... 439 D5.4 Bioptic Tissue Sampling Techniques and Needles ........................................................................................................ 440 D5.4.1 Fine-Needle Aspiration .................................................................................................................................... 440 D5.4.2 Trucut or Core Biopsy with Automated Biopsy Guns ...................................................................................... 440 D5.4.3 Coaxial Technique with Introducer Needle .................................................................................................... 441 References ................................................................................................................................................................................. 442
D5.1 INTRODUCTION A biopsy (in Greek bios = life and opsis = appearance) is a diagnostic procedure involving the removal and histopathologic examination of cells or tissues. Imaging modalities like computed tomography (CT) or magnetic resonance imaging (MRI) are powerful nondestructive techniques for producing cross-sectional images of the inside of the human body. Although these techniques are suitable for visualizing forensic relevant macropathologies like skeletal trauma, internal organ injuries, as well as pathologic gas collections, the actual obtained image resolution is not high enough for a reliable examination on the histological level. Promising steps in terms of a “virtual histology” with micro-CT and micro-MR have been taken [1], but these techniques are still areas of ongoing research and are not established in the forensic routine. Image-guided needle biopsy is a minimally invasive technique. It allows the collection of representative samples of organ tissue and body fluids (Figure D5.1) through small punctures. The harvested tissue is routinely processed and stained for histological and immunohistochemical examinations. A toxicological analysis of the obtained body fluids (i.e., blood, urine, cerebrospinal fluid) is also a diagnostic possibility. Needle biopsy in the postmortem setup is much easier to perform than in a clinical environment because of the absence of “moving targets” and the possibility of multiple biopsy attempts in the case of a misplaced needle tip. This method
has obvious benefits such as greater family acceptance due to the minimally invasive approach, decreased examination time, and, not least, the minimized risk of autopsy-related infections of the involved forensic pathologist.
D5.2 HISTORIC DEVELOPMENT D5.2.1 NON-IMAGE-GUIDED BIOPSY AND NEEDLE AUTOPSY The use of postmortem biopsy as a minimally invasive approach instead of body autopsy has a long history. Autopsies by needle—not image guided—have been used since the second half of the twentieth century. The first detailed report is that of Terry [2] in 1955—though he cited a previous description in a text on tropical disease in 1950—on needle sampling of tissues for the investigation of yellow fever and leishmaniasis in a South American population loathe to traditional autopsies [2]. Terry used a modification of an instrument designed for liver biopsy and indicated that most major organs could be sampled, including the brain. Even though the brain may be approached through the foramen magnum, the orbit, or the foramen oval, Terry proposed the approach through a small window in the skull above the region of interest. His personal series had a high success rate, with diagnoses made in 22 of 24 cases. Soon after Terry’s paper appeared, West and Chomet reported a series of 50 needle autopsies checked by subsequent 437
© 2009 by Taylor & Francis Group, LLC
438
The Virtopsy Approach
A
B
FIGURE D5.1 Minimally invasive sampling of heart blood with a puncture of the right ventricle under CT fluoroscopic control: (a) fluoroscopic picture; (b) coaxial needle with mounted syringe.
conventional autopsies [3]. The authors judged the procedure as being limited in its value, reporting discrepant diagnoses in 26 of their cases. The most common causes of discrepancy were myocardial infarction, pulmonary infarction, cerebrovascular accidents, heart failure, and gastrointestinal tract hemorrhages. In 1969, Wellmann reported a series of 394 consecutive needle autopsies, of which 37 had been limited to one or two organs at the request of the relatives [4]. In each of the remaining cases, attempts were made to sample the liver, heart, lung, and kidney, with success rates of 92%, 55%, 46%, and 34%, respectively. Like West and Chomet, Wellmann considered needle autopsy an inadequate substitute for conventional autopsy but still valuable when consent for full autopsy was not obtained. The 1983 study reported by Benbow [5] and the small series of interesting cases described by Underwood [6] in 1983 rekindled the interest to needle autopsies. An interesting aspect of the needle autopsies noted by Underwood was that the small-tissue samples tended to be more carefully scrutinized than the larger samples from conventional autopsy; a wider range of techniques and stains was applied to obtain as much information as possible out of a very limited sample. Although traditional autopsy can be carried out safely in HIV-positive subjects, some authors prefer to perform needle autopsies [7]. Baumgart et al. (1994) reported 16 cases where needle autopsy revised the cause of death in seven patients, including two previously undiagnosed cases of tuberculosis; several other unanticipated infectious diseases were also found [8]. An Australian study of 21 needle autopsies examining the liver in all cases but the kidney only in two, identified a cause of death in nine cases. In one of these, where death had been ascribed to ischemic heart disease, a large pulmonary thromboembolus was missed [9]. These were all coronial autopsies, apparently on subjects dying in community, and a pathologist guessing the cause of death prior to autopsy achieved 11 correct diagnoses on the bases of external examination and a history taken by the police.
© 2009 by Taylor & Francis Group, LLC
In 1996, Huston et al. [10] reported a prospective study of 20 consecutive cases examined by percutaneous needle biopsies before conventional autopsy. Liver and heart samples were recovered from all 20 cases, tissue of the lung from 18 (90%) and of the kidney from 16 cases (80%). The cause of death was confirmed in 67% of deceased. Needle sampling correlated with the complete autopsy in 87% of the additional major diagnoses and with equally pertinent negative results. Huston et al. also cultured the tissue specimens, which was a new technique in comparison with previous studies. Postmortem needle lung cultures correlated with the complete autopsy in 17 (85%) of 20 cases and 16 (80%) of 20 spleen cultures. Cultures of the brain (one case), cerebrospinal fluid (two cases), peritoneum (two cases), and serum (one case) correlated in 100% when compared with complete autopsy. In 1998 Aranda et al. [11] compared microbiological cultures obtained at autopsy with those received by fine-needle aspiration (FNA) puncture in 92 adult cases. They concluded that FNA puncture performed in the immediate postmortem period adds relevant microbiological information to the clinicopathologic picture and provides higher specificity than autopsy cultures. In a recent study of 150 needle autopsies of HIV-positive subjects, in 41 cases Guerra et al. [12] revealed diseases that had not been clinically suspected. They had clearly greater success than Wellmann [4], harvesting liver in 98%, lung in 93%, and kidney in 73% of cases, but they did not check their diagnoses by full autopsy. Recapitulating the previous studies, tissue sampling performed without any guidance is not of equal value in all cases compared to full autopsy. However, a number of advantages of this postmortem examination method were documented. Most of the authors practicing needle autopsies acknowledged the difficulty of obtaining samples from the desired organs or their parts, which made the diagnostic reliability of the method worse than that of traditional complete autopsy [3,6,8,12].
Biopsy
439
A
B
FIGURE D5.2 Fluoroscopic-guided biopsy of the liver (A) with the related histologic specimen (B).
D5.3 TARGETING PROCEDURES FOR CORRECT NEEDLE PLACEMENT D5.3.1 ULTRASOUND-GUIDED BIOPSY Needle placement is taken under sonographic control [13]. This method is suitable for the most parenchymal organs. Planning of the needle path is sometimes limited through system-dependent artifacts like from gas or bone. It is possible to perform the biopsy with a hands-free technique, where the needle is placed independently from the ultrasound (US) transducer, or with a specially designed probe that provides rigid needle guidance. The quality of the taken biopsies depends strongly on the experience of the examining person. Advantageously, there is no x-ray exposition of the examiner.
D5.3.2 MAGNETIC RESONANCE IMAGING MRI is an established tool in postmortem imaging [14–16], but to date there have been no attempts known to the authors to use this method for tissue sampling in a postmortem context. MRI provides superior anatomical information on the targeted soft tissue but at a slightly reduced spatial resolution compared with radiography. In contrast to radiography, MRI does not expose the examiner to ionizing radiation. Handling of ferromagnetic tools near an MRI scanner is impractical, so specially designed biopsy sets have to be used. MRI-guided interventions are an emerging field in clinical radiology [17–19]; alone the possibility of an automated (magnetic field driven) needle and catheter positioning is an exciting forecast of the interventional potentials of this imaging modality.
D5.3.3 CT-GUIDED BIOPSY The high spatial resolution visualization of the inside of the body without superimposition of structures and the possibility of a nearly real-time needle control in CT fluoroscopy are the main advantages in biopsy planning with CT. There are several techniques for performing CT-guided biopsy.
© 2009 by Taylor & Francis Group, LLC
D5.3.3.1 Classic Step-and-Shot Technique Planning of the needle approach is undertaken with previously acquired CT cross-sections, and control of the needle position is obtained with subsequent scans of the region of interest. The examining person can avoid radiation exposure by leaving the room during the control scans. Needle insertion should be strictly in the scan plane of the CT in order to provide a proper display of the biopsy needle. Needle angulation in the Z (CT couch)-axis is possible by tilting the CT gantry in the needed position. Taking multiple biopsies from different organs makes this technique quite time-consuming. D5.3.3.2 CT Fluoroscopy The recent development of rapid reconstruction of CT images has enabled scanners the added benefit of performing “realtime” image-guided procedures [20–22]. The main advantage of this technique is the markedly decreased procedure time compared with use of conventional CT guidance. Planning of the needle approach is also undertaken with previously acquired CT data, but placement of the needle itself is performed under real-time control with the body in the CT gantry (Figure D5.2). The examiner is standing next to the CT couch with direct control of the needle placement and instant access to the fluoroscopy functions of the scanner (Figure D5.3). The real-time simultaneous representation of the target slice and the adjacent superior and inferior slices allows an immediate identification of the needle tip and direction. Due to the x-ray exposition of the examination, personal care has to be taken in terms of radiation protection. D5.3.3.3 Navigated Biopsy with Robotic Assistance A robotic-assisted navigated biopsy device using the guidance of CT (Figure D5.4) is of great value for several reasons: (1) It provides very stable needle guidance, even for angulated approaches, (2) it allows access to lesions when the width of the CT gantry would limit the access for a biopsy needle, and (3) it provides superior, user-independent accuracy of needle placement. The examiner has to label the target on the previously obtained CT images, and the system computes the
440
The Virtopsy Approach
FIGURE D5.3 CT scanner (SOMATOM 6, Siemens Medical Solutions, Forchheim) equipped with a fluoroscopy package. Control monitor in the left upper edge, joystick for manual table movement on the right side of the CT couch, and foot switch for triggering of the fluoroscopy function.
biopsy trajectory and performs the biopsy autonomously after the clearance of the user. Positions of the body and the biopsy device are permanently monitored by a navigation system. The function of this system is similar to the global positioning systems (GPSs) used in cars. Interventional robotic systems
are a young and evolving field [23,24]. The implementation of such a system in the Virtobot concept [25] is one of the next projects in the Virtopsy Group.
D5.4 BIOPTIC TISSUE SAMPLING TECHNIQUES AND NEEDLES D5.4.1 FINE-NEEDLE ASPIRATION FNA is a percutaneous procedure that uses a fine-gauge needle (22 G). Typically, a Chiba needle (18–22 G) with an inner trocar is used. The FNA allows only cytological and microbiological diagnostics due to the minimal removal of cell clusters. The thin caliber and the length of the needle aggravate dynamic effects in the soft tissue, leading to a higher misguiding of the inserted FNA needles. For these reasons this method is not suitable for a comprehensive postmortem histopathological examination [26].
D5.4.2 TRUCUT OR CORE BIOPSY WITH AUTOMATED BIOPSY GUNS FIGURE D5.4 Computer model of biopsy system with robotic assistance, planned by the Virtopsy Group in Bern.
© 2009 by Taylor & Francis Group, LLC
The use of automated biopsy devices, or biopsy guns, has become commonplace in clinical radiology. There are multiple similar working systems on the market. The automated
Biopsy
441
FIGURE D5.5 Automatic biopsy gun (Bard Magnum) with installed Trucut needle. Note the arrow-shaped marker shield for passive registration with navigated biopsy system.
biopsy devices are easy to use, reliable, and generally provide a comparable specimen for histopathologic analysis [27–29]. The Virtopsy Group uses an automatic biopsy gun (Bard Magnum®) (Figure D5.5). Trucut biopsy needles are designed for the capture of high-quality tissue samples by removing a hemicylinder from the targeted tissue volume. The most typical caliber range is from 14 to 18 G. The principle of Trucut biopsy is illustrated in Figure D5.6. The first
A
attempts to implement postmortem biopsy in the forensic routine showed promising results [30].
D5.4.3 COAXIAL TECHNIQUE WITH INTRODUCER NEEDLE In the coaxial technique, a hollow needle with an inner trocar is used. The biopsy is taken through the lumen of the hollow needle after removing the inner trocar (Figure D5.7). The major advantage of this technique is the possibility of taking multiple tissue samples from the same target volume without the need for a new puncture for every biopsy. In the CT gantry there is often not enough space to place a long Trucut needle with the connected biopsy gun; therefore, the handling of the shorter introducer is much easier to perform in a significantly shorter time.
B
C
D
FIGURE D5.6 Sequence of Trucut biopsy: (A) Placing of the needle tip short before the volume of interest. (B) Inner trocar with hemicylindric notch shoots forward with high velocity (up to 12 m/s). (C,D) Hollow cannula shots over the inner trocar, separating the tissue in the notch from the surrounding tissue.
© 2009 by Taylor & Francis Group, LLC
FIGURE D5.7 Trucut needle with installed biopsy pistol is inserted in the introducer needle (after removing the blind ending inner trocar).
442
REFERENCES 1. Thali MJ, Dirnhofer R, Becker R, Oliver W, Potter K. 2004. Is “virtual histology” the next step after the “virtual autopsy”? Magnetic resonance microscopy in forensic medicine. Magn Reson Imaging 22:1131–38. 2. Terry R. 1955. Needle necropsy. J Clin Pathol 8(1):38–41. 3. West M, Chomet B. 1957. An evaluation of needle necropsies. Am J Med Sci 234(5):554–60. 4. Wellmann K. 1969. The needle autopsy. A retrospective evaluation of 394 consecutive cases. Am J Clin Pathol 52(4):441–44. 5. Benbow EW. 1974. The needle necropsy. Br Med J (Clin Res Ed) 1983;286(6382). 6. Underwood JC, Slater DN, Parsons MA. 1983. The needle necropsy. Br Med J (Clin Res Ed) 286(6378):1632–34. 7. Lucas SB. 1993. HIV and the necropsy. J Clin Pathol 46(12):1071–75. 8. Baumgart KW, Cook M, Quin J, Painter D, Gatenby PA, Garsia RJ. 1994. The limited (needle biopsy) autopsy and the acquired immunodeficiency syndrome. Pathology 26(2):141–43. 9. Foroudi F, Cheung K, Duflou J. 1995. A comparison of the needle biopsy post mortem with the conventional autopsy. Pathology 27(1):79–82. 10. Huston BM, Malouf NN, Azar HA. 1996. Percutaneous needle autopsy sampling. Mod Pathol 9(12):1101–07. 11. Aranda M, Marti C, Bernet M, Gudiol F, Pujol R. 1998. Diagnostic utility of postmortem fine-needle aspiration cultures. Arch Pathol Lab Med 122(7):650–55. 12. Guerra I, Ortiz E, Portu J, Atares B, Aldamiz-Etxebarria M, De Pablos M. 2001. Value of limited necropsy in HIVpositive patients. Pathol Res Pract 197(3):165–68. 13. Fariña J, Millana C, Fdez-Aceñero JM, et al. 2002. Ultrasonographic autopsy (echopsy): a new autopsy technique. Virchows Arch 440:635–39. 14. Yen K, Vock P, Tiefenthaler B, et al. 2004.Virtopsy: forensic traumatology of the subcutaneous fatty tissue; multislice computed tomography (MSCT) and magnetic resonance imaging (MRI) as diagnostic tools. J Forensic Sci 49:799–806. 15. Griffiths PD, Paley MNJ, Whitby EH. 2005. Post-mortem MRI as an adjunct to fetal or neonatal autopsy. Lancet 365:1271–73. 16. Jackowski C, Christe A, Sonnenschein M, Aghayev E, Thali MJ. 2006. Postmortem unenhanced magnetic resonance imaging of myocardial infarction in correlation to histological infarction age characterization. Eur Heart J 27:2459–67. 17. Saborowski O, Saeed M. 2007. An overview on the advances in cardiovascular interventional MR imaging. Magn Reson Mater Phy 20:117–27.
© 2009 by Taylor & Francis Group, LLC
The Virtopsy Approach
18. Fischer GS, Dyer E, Csoma C, Deguet A, Fichtinger G. 2007. Validation system of MR image overlay and other needle insertion techniques. Stud Health Technol Inform 125:130–35. 19. Khamene A, Wacker F, Vogt S, et al. 2003. An Augmented Reality system for MRI-guided needle biopsies. Stud Health Technol Inform 94:151–57. 20. Bissoli E, Bison L, Gioulis E, Chisena C, Fabbris R. 2003. [Multislice CT fluoroscopy: technical principles, clinical applications and dosimetry]. Radiol Med (Torino) 106:201–12. 21. Gianfelice D, Lepanto L, Perreault P, Chartrand-Lefebvre C, Milette PC. 2000. Value of CT fluoroscopy for percutaneous biopsy procedures. J Vasc Interv Radiol 11:879–84. 22. Paulson EK, Sheafor DH, Enterline DS, McAdams HP, Yoshizumi TT. 2001. CT fluoroscopy-guided interventional procedures: techniques and radiation dose to radiologists. Radiology 220:161–67. 23. Solomon SB, Patriciu A, Bohlman ME, Kavoussi LR, Stoianovici D. 2002. Robotically driven interventions: a method of using CT fluoroscopy without radiation exposure to the physician. Radiology 225:277–82. 24. Cleary K, Watson V, Lindisch D, et al. 2005. Precision placement of instruments for minimally invasive procedures using a “needle driver” robot. Int J Med Robot 1:40–47. 25. Dirnhofer R, Jackowski C, Vock P, Potter K, Thali MJ. 2006. VIRTOPSY: minimally invasive, imaging-guided virtual autopsy. Radiographics 26:1305–33. 26. Hopper KD, Abendroth CS, Sturtz KW, Matthews YL, Hartzel JS, Potok PS. 1995. CT percutaneous biopsy guns: comparison of end-cut and side-notch devices in cadaveric specimens. Am J Roentgenol 164(1):195–99. 27. De Man RA, van Buuren HR, Hop WCJ. 2004. A randomised study on the efficacy and safety of an automated Tru-Cut needle for percutaneous liver biopsy. Neth J Med 62:441–45. 28. Hopper KD, Abendroth CS, Sturtz KW, Matthews YL, Shirk SJ, Stevens LA. 1993. Blinded comparison of biopsy needles and automated devices in vitro: 1. Biopsy of diffuse hepatic disease. AJR Am J Roentgenol 161(6):1293–97. 29. Hopper KD, Abendroth CS, Sturtz KW, Matthews YL, Shirk SJ, Stevens LA. 1993. Blinded comparison of biopsy needles and automated devices in vitro: 2. Biopsy of medical renal disease. AJR Am J Roentgenol 161(6):1299–301. 30. Aghayev E, Thali MJ, Sonnenschein M, Jackowski C, Dirnhofer R, Vock P. 2007. Post-mortem tissue sampling using computed tomography guidance. Forensic Sci Int 166:199–203.
D6
Postmortem Angiography
CONTENTS D6.1 Postmortem Angiography: Historical Review and Overview of Former and Current Techniques .............................. 443 D6.1.1 Development of Postmortem Angiography ................................................................................................... 444 D6.1.2 Overview of Postmortem Angiographic Methods ......................................................................................... 444 D6.1.2.1 Corpuscular Solutions ................................................................................................................ 444 D6.1.2.2 Oily Liquids ................................................................................................................................ 444 D6.1.2.3 Hydrosoluble Liquids.................................................................................................................. 445 D6.1.2.4 Casts ........................................................................................................................................... 447 D6.1.2.5 Other Injecting Masses ............................................................................................................... 447 D6.1.3 Application Techniques ................................................................................................................................. 448 D6.1.3.1 Single-Organ Angiography ........................................................................................................ 448 D6.1.3.2 Whole-Body Angiography ......................................................................................................... 449 D6.1.4 Imaging .......................................................................................................................................................... 449 D6.1.5 References ...................................................................................................................................................... 449 D6.2 Post-Mortem Angiography after Vascular Perfusion with Diesel Oil and a Lipophilic Contrast Agent ...................... 451 D6.2.1 Methodological Concept ................................................................................................................................ 451 D6.2.1.1 Perfusate ..................................................................................................................................... 451 D6.2.1.2 Pump........................................................................................................................................... 451 D6.2.1.3 Contrast-Agent ............................................................................................................................ 452 D6.2.2 Two-Step Method .......................................................................................................................................... 452 D6.2.2.1 Establishing a Postmortem Circulation ...................................................................................... 452 D6.2.2.2 Angiography ............................................................................................................................... 453 D6.2.3 Results of the Technique ................................................................................................................................ 453 D6.2.4 References ...................................................................................................................................................... 456 D6.3 Postmortem Minimally Invasive Angiography ............................................................................................................. 456 D6.3.1 Ex Vivo Porcine Model ................................................................................................................................. 457 D6.3.2 Human Corpse Model.................................................................................................................................... 459 D6.3.3 The Method and the Existing Literature ...................................................................................................... 461 D6.3.4 The Injection Pressure ................................................................................................................................... 462 D6.3.5 The Injected Contrast-Agent Volume ............................................................................................................ 464 D6.3.6 The Contrast Agents ...................................................................................................................................... 464 D6.3.7 But How to Overcome the Extravasation of Contrast Agent? ....................................................................... 465 D6.3.8 High-Viscosity Angiography on a Human Corpse ........................................................................................ 467 D6.3.9 General Thoughts and Perspectives .............................................................................................................. 468 D6.3.10 Acknowledgments ......................................................................................................................................... 469 D6.3.11 References ...................................................................................................................................................... 470
D6.1
POSTMORTEM ANGIOGRAPHY: HISTORICAL REVIEW AND OVERVIEW OF FORMER AND CURRENT TECHNIQUES Silke Grabherr and Richard Dirnhofer
In classical autopsy, demonstration of the vascular system can be a problematic point. While it is easy to show the great vessels like the aorta, the iliac vessels, or the main vessels of extremities and
organs, smaller ones are difficult to examine. Preparing small vasculature is a time-consuming and sometimes impossible exercise. Also, with the use of non-contrast-enhanced (native) multislice computed tomography (MSCT) and magnetic resonance imaging (MRI) the great vessels can be demonstrated, but the smaller ones are not visible because the resolution of the techniques is too low to demonstrate smaller vessels in detail. Like clinical in vivo angiography, postmortem angiography is a tool for performing vascular diagnosis. Principally,
443 © 2009 by Taylor & Francis Group, LLC
444
The Virtopsy Approach
four types of investigations exist in which postmortem angiography is helpful: r Organ-specific analysis of vascular pattern and anatomical variations r Pathological and physiological observations r Changes induced by unnatural causes r Experimental testing of novel contrast agents While older studies were performed to learn about the anatomy of the vascular system, modern studies are mostly indicated to analyze pathological and physiological changes and, especially in forensic medicine, to detect changes induced by unnatural causes like traumatic ruptures of vessels.
D6.1.1 DEVELOPMENT OF POSTMORTEM ANGIOGRAPHY The vascular system has always been a field of research in the history of human medicine. Various substances have been injected into vessels, and many methods have been performed to answer these morphological questions. The first reported preparations of hollow anatomical structures were done by Leonardo da Vinci and Jakobus Berengius around 1500 [1]. They produced casts of the heart chambers and cerebral ventricles that were made of wax and were cleaned from the surrounding tissue by maggots. This first maceration technique was replaced by chemical maceration methods in the following years. At the end of the 18th century another method to demonstrate the injected vessels was developed. It was a kind of lightning technique that was described by Virchow [2] in 1857, using an increasing concentration of alcohol. After the discovery of x-rays, radioopaque material was added to the injecting masses, and angiography was born. The first to report such a postmortem angiography was Baumgarten, who injected radioopaque material into the coronary arteries of isolated human hearts in 1899 [1]. The use of contrast agents increased rapidly in the first years of the 20th century. The number of different methods and injecting substances reached its peak in the first half of the 20th century. Detailed lists of the performed methods can be found in Schoenmackers [1] and Grabherr [3]. Today, there are only two common methods used to perform postmortem angiography, such as the use of barium sulfate and silicone rubber. The other techniques disappeared from publications and seem to be buried in oblivion. As the use of postmortem radiographic diagnostics has increased rapidly in forensic medicine and pathology, the performance of postmortem angiography is experiencing a revival and has led to further evolution of common methods and to the development of new techniques.
D6.1.2 OVERVIEW OF POSTMORTEM ANGIOGRAPHIC METHODS When subdividing the different techniques according to the nature of the injection material, five different types can be found: (1) corpuscular solutions; (2) oily liquids; (3) hydrosoluble liquids; (4) casts; and (5) miscellaneous (Table D6.1.1).
© 2009 by Taylor & Francis Group, LLC
D6.1.2.1
Corpuscular Solutions
Materials from this group were used most frequently for postmortem angiography. They consist of a corpuscular radioopaque medium that is soluble in a liquid—mostly in water. Corpuscular solutions have to be mixed shortly before injecting them. The first medium that was reported to be used for angiography was Menninge. This powder of lead oxide with its typical red color was first added to an injecting mass by Jamin in 1907 [1] and later by Mitaya [4] (Figure D6.1.1). While the application of mennige decreased, barium sulfate, another radioopaque medium, was becoming more and more common. Stoeter used barium sulfate to demonstrate the vascular system of human embryos and fetuses in 1976 [5]. His images of the whole body are impressive (Figure D6.1.2). Using a solution of water and finer-grained barium sulfate, it was possible to investigate smaller vessels and also capillaries. This contrast medium, called Micropaque, allowed a new technique of postmortem angiography called microangiography [6,7]. If the corpuscular substance is dissolved in a warm solution of gelatin or agar, the injecting mass hardens inside the vessels after cooling. This method is very practicable if histological examinations follow the angiography. Because of the higher viscosity, these solutions cannot reach the capillaries and smaller arterioles. The warm solution of barium in gelatin or agar seems to be the most frequently used injection material. It can be found in literature from 1924 [8] through the present [9]. A pioneer in using this injection group was Schlesinger [10], whose technique of the “unrolled” heart was published in 1938 and was modified and republished by many other authors [11,12]. Apart from the barium solution, other corpuscular solutions have been used less frequently (Table D6.1.1). All injection materials with corpuscular particles are characterized by attributes that were described by Frik and Persch in 1969 in a study that compared water-soluble, oily, and corpuscular contrast agents [13]. The authors noticed that corpuscular materials have the ability to displace postmortem clots and to produce a high contrast in angiography. Marxen [14] described artifacts caused by precipitation of the corpuscular particles during or before imaging. Advantages to this technique are its demonstration of microcirculation when dissolved in water, the ability to flush out postmortem clots, high radioopacity, and its practicability for histological studies when solved in gelatin or agar. Disadvantages include extravasations when dissolved in water, no demonstration of microcirculation when dissolved in gelatin or agar, and precipitation-induced artifacts. D6.1.2.2
Oily Liquids
Applications of oily contrast agents for postmortem angiography were reported less often than the use of hydrophil liquids and corpuscular solutions. Parade [15] was the first to inject a commercially available Jodipin solution to demonstrate the coronary arteries in 1933. The number of lipophilic contrast agents is rare. For postmortem angiography Dinosil [16] and Lipiodol Ultrafluide [17] have been used. Oily liquids have also been used for postmortem because of their ability to stay intravascular without penetrating the
Postmortem Angiography
445
TABLE D6.1.1 Overview of Injection Materials for Postmortem Angiography Group
Material (References)
1. Corpuscular preparations
(a) Prepared in gelatin or agar: Barium sulfate in agar or gelatine [10,11] Potassium iodine in gelatin or agar [48] Menninge in gelatin [4] Lead sulfate in gelatin or agar [11] (b) Prepared in watery solution Barium sulfate in water [4,5] Microopaque [6,14] Bismuth chloride in water [38] Potassium iodine and Karo corn syrup [48] Jodipin solution [15] Dinosil [16] Lipiodol Ultrafluide [18,21] Mixture of diesel oil and paraffin oil [20] Diesel oil and Lipiodol Ultrafluide [18]
2. Oily liquids
3. Hydrosoluble preparations
4. Casts
5. Miscellaneous
Formalin-containing dyes [27] Cardiografin [22] Hypaque [23] Gastrograffin [25] Telebrix Gastro [26]
Various metals, such as lead, bismuth, cadmium, rose metal, Wood’s metal [1] Nylon compounds [50] Neoprene latex [33] Vinyl (51) Polyester resins (52) Silicone rubber and lead oxide [30] Mercox [45] Microfil [44] Polyurethane-based compounds [53] Special mixtures [54–56]
Advantages
Disadvantages The microcirculation is not revealed
Ability to flush out postmortem clots
Ability to flush out postmortem clots Visualization of microcirculation Ability to flush out postmortem clots No extravasation Retained intravascularly for long periods The death-to-injection and the injection-to-imaging intervals can be extended Easy to inject
Suitable for single-organ studies Artifacts are rare
Mixture-dependent
Disadvantages that have been discussed in the literature but that are not proven are marked with“?”.
surrounding tissue [18,19]. For example, Barmeyer [20] used a mixture of diesel oil and paraffin oil for a perfusion of the coronary arteries. The method of establishing a postmortem circulation by the use of diesel oil or paraffin oil as a perfusat will be shown in detail later. A list of different types of oils suited for injection into the vascular system can be found in Schoenmackers [1]. The abilities of oily liquids for angiography were well described by Pfeifer in 1947 [21] (Figure D6.1.3). Like corpuscular solutions, they are able to wash out postmortem clots and to provide a strong contrast in angiography. One negative aspect of oily liquids was described by Schoenmackers (1). He mentioned that they can infiltrate aggrieved areas of the vascular wall and unhinge cholesterol and lipids. Depending on the
© 2009 by Taylor & Francis Group, LLC
Precipitation can occur before imaging Extravasation Infiltration of damaged vascular walls (?) Extraction of lipids from damaged walls (?) Demonstration of the microcirculation depends on the viscosity of the medium Rapidly penetrate the surrounding tissue Edema Low contrast Vessels appear thinner than in reality (?) Artifacts generated by postmortem clots Not practicable for whole-body injection Shrinkage during hardening Cannot be flushed out
Mixture-dependent Preparation procedure is usually complex
viscosity, arterioles or capillaries cannot be perfused and demonstrated because the oil leads to microembolization [18]. Advantages of this method include long intravascular retention time, no extravasation; ability to be used in a late postmortem stage, ability to flush out postmortem clots, and ability to extend the interval between injection and imaging. Disadvantages are speculated infiltration of aggrieved vascular walls with unhinging of lipids from the vascular wall, and viscosity-depending demonstration of microcirculation. D6.1.2.3
Hydrosoluble Liquids
For in vivo angiography the use of these liquids is essential, but for postmortem studies they are not that common.
446
The Virtopsy Approach
FIGURE D6.1.1 Vascular system of the lung, as revealed after injecting a solution of menninge. (From Mitaya, S., Virchows Arch Path Anat 304:608–24, 1939. With permission.)
Different contrast agents that are mostly used in clinical in vivo examinations were also applied postmortem, such as Cardiografin [22], Hypaque [23], Cloropaque [24], Gastrograffin [25], and Telebrix Gastro [26], or simply formalin with added dyes [27] or hydrosoluble contrast agents that have not been defined [28] (Figure D6.1.4). The first reported application of hydrosoluble liquids for postmortem purposes was in 1866 by Chrzonszczewsky [29]. Hydrosoluble agents have the advantage of being very easy to inject, and their application is not time consuming. The negative aspect is the rapid penetration through the vessel’s wall that causes edema of the surrounding tissue. However, as described by Zapata [17] and by Frik and Persch [13], the quality of images obtained with water-soluble liquids is inferior. Frik and Persch [13] also reported that the observed
© 2009 by Taylor & Francis Group, LLC
vessel’s diameter varies in angiograms performed with injections of different liquids. While the diameter does not show a significant difference between oily and corpuscular liquids, vessels injected with hydrosoluble liquids appeared thinner at imaging. The reason for this phenomenon was thought to be that water-soluble contrast agents are transported in an axial direction inside the vessels and do not fill the whole lumina. The authors also mentioned that perhaps the oily and corpuscular solutions lead to a dilatation of the perfused vessels. An advantage to this method is that it is easy and fast to apply. Disadvantages include rapid penetration into surrounding tissue with edematization, low contrast, speculation of letting vessels appear thinner than they are, and artifacts because of postmortem clots.
Postmortem Angiography
447
FIGURE D6.1.3 Single-organ angiograph of a kidney, which was prepared by injecting the liposoluble contrast agent Lipiodol Ultrafluide. (From Pfeifer, K. J., U. Klein, C. H. Chaussy, et al., Fortschr Röntegenstr 121:472–76, 1974. With permission.)
FIGURE D6.1.2 Whole-body angiograph of a human fetus, which was prepared by perfusing a solution of barium sulfate. (From Stoeter, P. and K. Voigt, Fortschr Geb Roentengstr 124:558–64, 1976. With permission.)
D6.1.2.4
Casts
This oldest form of visualizing the vascular system has its roots in the 15th century, as described already. Leonardo da Vinci was the first person known to use this technique [1]. The requirements for injection materials were different at this time, when x-rays were unknown. The idea of these methods was to inject a liquid or a mass into the vascular system that was able to harden; following this, the surrounding tissue was removed by maceration techniques to visualize the hardened injection mass. Around 1700, Bidloo [1] adopted different metals with a low melting point as injection material. Some examples of the many other injected materials can be found in Table D6.1.1 (Figure D6.1.5). The casting method, first presented in 1987 by SegerbergKottinen [30] using a silicone rubber with lead oxide, is the most practiced method in forensic medicine. This technique
© 2009 by Taylor & Francis Group, LLC
allows the demonstration of blood vessels even with a diameter of 0.1 mm. The application of casting materials is not always easy to perform, but their frequent use shows their success in postmortem angiography. While the early casting materials were complex mixtures of many ingredients, the mixtures today are commercially available and easy to prepare. Advantages are that artifacts are rare and that the method is good for single-organ studies. Disadvantages include shrinkage after hardening, inability to be flushed out after application, and impracticability for whole-body angiography. D6.1.2.5
Other Injecting Masses
There are some injecting masses reported that cannot be divided into the four groups described above. Most of them consist of different ingredients from groups described already, and have been developed to answer specific questions. Some authors have been very imaginative by mixing their own creations. Numerous formulas existed that allowed preparing injection masses in different colors and viscosities and for different purposes. Some of them can be found in Schoenmackers [1] and in Table D6.1.1 (Figure D6.1.6). The abilities of these materials depend on their composition, but most of them are very time consuming to prepare.
448
The Virtopsy Approach
FIGURE D6.1.4 Whole-body angiograph, as revealed after the injection of an undefined water-soluble contrast agent. (From Foote, G.A., Wilson, A.J., and Sleword, J.H. Br. J. Radiol. 51:351–356, 1978. With permission.)
FIGURE D6.1.6 Angiograph of the testicular vessels, which were injected with a mixture of menninge, terpentine oil, and vaseline. (From Cocchetti, E. and I. Donini, Ateneo Parmese 25:318–36, 1954. With permission.)
D6.1.3 APPLICATION TECHNIQUES In addition to the existence of many different injecting materials, the technique of applying these materials can vary. Studies can be subdivided into single-organ angiographies or whole-body angiographies. D6.1.3.1 FIGURE D6.1.5 Vinylite cast of the coronary arteries prepared by Stern in 1954 (51). (Adapted from Stern, H., E. R. Ranzenhofer, and A. A. Liebow, Lab Invest 30:337–47, 1954. With permission.)
© 2009 by Taylor & Francis Group, LLC
Single-Organ Angiography
Principally, the organs can be injected in situ [31] or after organ removal. While in situ injections are rare, the approach of removing the organ before contrast-agent injection is
Postmortem Angiography
widely adopted. The best described technique is that of Schlesinger [10], where the freshly removed heart is warmed up and perfused in a water bath. The coronary arteries are cannulated and washed with warm physiologic salt solution that is injected at 150 mmHg pressure. After this, warm lead agar or barium gelatin mass is injected at the same pressure. Other techniques do not differ much from Schlesinger’s. Generally, the organs are put in a warm water bath and then the contrast mass is injected, especially if it is a mass of high viscosity. If the viscosity of the contrast agent is low (e.g., by the use of a hydrosoluble liquid), the organ is injected without the use of a water bath [26]. D6.1.3.1.1 Organ Treatment Prior to Contrast-Agent Injection Often described in the literature is the importance of rinsing the vessels to wash out blood and postmortem clots before injecting the contrast agent, to avoid artifacts caused by them. Spalteholz, who was one of the pioneers in postmortem vascular injections, perfused the coronary arteries with a warmed-up salt solution of 38°C [32]. This method was copied by others. Some authors declared that the organ should be fixated before the injection of contrast agents [33], and even decalcified [34]. In comparison with rinsing out the vascular system before angiography, other authors [35] achieved good results by injecting the contrast medium directly without performing a precursory perfusion. D6.1.3.1.2 Injection Pressure The optimal pressure for injection is a matter of controversy. Advised pressures are 40–60 mmHg [1], 100 mmHg [36], 120–180 mmHg (37) and 220 mmHg (38). To reach the designated injection pressure, different techniques were used, such as gravity [39]. Different pressure regulators [40], roller pumps [18], and a modified heart lung machine [41] were also used. For other authors, careful manual injection was the method least likely to damage the vessels [42]. D6.1.3.2
Whole-Body Angiography
The small number of reported whole-body perfusions was performed on animal and human fetuses, embryos, and newborns [5,43] shortly after death. Mostly the applied techniques were similar to those used for single-organ studies. The method described by Stoeter [5] is an exception. These investigators proposed radiologically controlled discontinuous injection of the contrast agent with angiography being performed in the intervening intervals to observe the filling of the vascular system. The whole-body angiographic methods of Grabherr [18] and Jackowski [25] are described in detail in the following section.
D6.1.4 IMAGING To inspect the filled vessels, the following tools have been implemented:
© 2009 by Taylor & Francis Group, LLC
449
r Macroscopic and microscopic inspection of threedimensional casts after maceration of the surrounding tissue r Conventional x-radiography r Xeroradiography r CT r MRI r Micro-CT r Scanning electron microscopy Using MSCT as the imaging technique seems to be optimal because it allows the vessels to be examined from different points of view and in a two- and three-dimensional manner. A whole-body angiography performed with MSCT and reconstructed as a 3-D model can give an easy overview of the entire vascular system that demonstrates a bleeding or a breakup of blood support in an easily understandable way, even for medical laypersons. Modern postmortem angiography is therefore mostly performed by the use of CT. Analagous to this development, implementation of micro-CT [44] scanners for microangiography has already started beside the common use of scanning electron microscopy [45] in this field. While conventional radiography is mentioned often in the literature, other imaging techniques such as xeroradiography [45] and MRI imaging [47] are rarely described to be used for postmortem angiography. Macroscopic and microscopic inspection of three-dimensional cast after maceration of the surrounding tissue does not play an important role in modern investigations, but in the early pioneering days of angiography these were the only ways to investigate the filled vessels.
D6.1.5 REFERENCES 1. Schoenmackers, J. 1960. Technik der postmortalen Angiographie mit Berücksichtigung verwandter Methoden postmortaler Gefäßdarstellung. Ergebn allgem Pathol Anat 39:53–151. 2. Virchow, R. 1857. Einige Bemerkungen über die Circulations-verhältnisse in den Nieren. Virchows Arch Path Anat 12:310–25. 3. Grabherr, S., V. Djonov, K. Yen, M. J. Thali, and R. Dirnhofer. 2007. Post-mortem angiography: A review of former and current methods. AJR 188:832–38. 4. Mitaya, S. 1939. Aufbau und Gestalt der peripheren arteriellen Strombahn des kleinen Kreislaufs. Virchows Arch Path Anat 304:608–24. 5. Stoeter, P. and K. Voigt. 1976. Radiological examination of embryonal and fetal vessels. Technique and method of prenatal, post-mortem angiography in different stages of gestation. Fortschr Geb Roentengstr 124:558–64. 6. Dor, P. and G. Salamon. 1970. The arterioles and capillaries of the brain stem and cerebellum: A microangiographic study. Neuroradiology 1:27–29. 7. Hassler, O. 1966. Deep cerebral venous system in man. A microangiographic study on its areas of drainage and its anastomoses with the superficial cerebral veins. Neurology 16:504–11. 8. Hinman, F. and D. M. Morison. 1924. Comparative study of circulatory changes in hydronephrosis, caseo-cavernous tuberculosis, and polycystic kidney. J Urol 11:131–41.
450
9. Rozenberg, V. D. and L. M. Nepomnyshchikh. 2002. Pathomorphology of myocardial bridges and their role in the pathogenesis of coronary disease. Bull Exp Biol Med 134:593–96. 10. Schlesinger, M. J. 1938. An injection plus dissection study of coronary artery occlusions and anastomosis. Am Heart J 15:528–68. 11. Rodriguez, F. L. and L. Reiner. 1965. Postmortem angiographic studies on the coronary arterial circulation. Am Heart J 70:348–64. 12. Heard, B. E. 1976. Pathology of hearts after aortocoronary saphenous vein bypass grafting for coronary artery disease, studied by post-mortem angiography. Br Heart J 38:838–59. 13. Frik, W. and W. F. Persch. 1969. Der Einfluß des Kontrastmitteltyps auf das Arterienkaliber om der experimentellen Angiographie. Fortschr Röntgenstr 11:620–29. 14. Marxen, M., M. M. Thornton, C. B. Chiarot, et al. 2004. MicroCT scanner performance and considerations for vascular specimen imaging. Med Phys 31:305–13. 15. Parade, G. W. 1933. Coronardarstellung. Verh Dtsch Ges Inn Med 45:216–20. 16. G. S. Melnick, N. Tuna, and M. J. Gilson. 1963. Postmortem coronary arteriogram. A correlation with electrocardiographic and anatomic findings. Angiology 14:252–59. 17. Van der Straeten, P. P. 1955. La coronarographie post mortem de l’homme age. Acta Cardiol 10:15–43. 18. Grabherr, S., V. Djonov, A. Friess, et al. 2006. Post-mortem angiography after vascular perfusion with diesel oil and a lipophilic contrast agent. AJR 187:W515–23. 19. Zapata, M. G., M. Alcaraz, and A. Luna. 1989. Study of postmortem blood circulation. Z Rechtsmed 103:27–32. 20. Barmeyer, J. 1968. Postmortale Koronarangiographie und Perfusion normaler und pathologisch veränderter Herzen, Messung der Durchflusskapazität interkoronarer Anastomosen. Beitr Pathol Anat 137:373–90. 21. Pfeifer, K. J., U. Klein, C. H. Chaussy, et al. 1974. Postmortale Nierenvergrößerungsangiographie mit fettlöslichem Kontrastmittel. Fortschr Röntegenstr 121:472–76. 22. Eusterman, J. H., R. W. P. Achor, O. W. Kincaid, and A. L. Brown. 1962. Artherosclerotic disease of the coronary arteries. A pathologic-radiologic correlative study. Circulation 26:1288-95. 23. McNamara, J. J., M. A. Molot, J. F. Stremple, and R. T. Cutting. 1971. Coronary artery disease in combat casualties in Vietnam. J Am Med Ass 216:1185–87. 24. Vesterby, A. 1981. Postmortem coronary angiography and histological investigation of the conduction system of the heart in sudden unexpected death due to coronary heart disease. Acta Pathol Microbiol Scand 89:157–63. 25. Smith, M., D. E. Trummel, M. Dolz, and S. J. Cina. 1999. A simplified method for postmortem coronary angiography using gastrograffin. Arch Pathol Lab Med 123:885–88. 26. Jackowski, C., M. Sonnenschein, M. Thali, et al. 2005. Virtopsy: postmortem minimally invasive angiography using cross section techniques—implementation and preliminary results. J Forensic Sci 50:1175–86. 27. Neumann, R. 1939. Die Cardiaorta als Organ und ihr Verhalten bei Coronarsklerose. Virchows Arch path Anat 303:1. 28. Foote, G. A., Wilson, A. J., and Steward, J. J. 1978. Perinatal post-mortem radiography experience with 2500 cases. Br. J. Radiol. 51:351–356. 29. Chrzonszczewsky, N. 1866. Zur Anatomie und Physiologie der Leber. Virchows Arch path Anat 35:153.
© 2009 by Taylor & Francis Group, LLC
The Virtopsy Approach
30. Segerberg-Kottinen, M. 1987. Demonstration of esophageal varices postmortem by gastroeosophageal phlebography. J Forensic Sci 32:703–10. 31. Schlichter, J. and H. Harris. 1949. The vascularisation of the aorta. Am J Med Sci 218:610–15. 32. Spalteholz, W. 1907. Anatomischer Teil. Dtsch Med Wochenschr 20:792–95. 33. Smith, J. R. and M. J. Henry. 1945. Neoprene latex demonstration of the coronary arteries. J Lab & Clin Med 30:462–66. 34. Trask, N., R. M. Califf, M. J. Conley, et al. 1984. Accuracy and interobserver variability of coronary cineangiography: a comparison with postmortem evaluation. J Am Coll Cardiol 3:1145–54. 35. Vesterby, A. 1981. Postmortem coronary angiography and histological investigation of the conduction system of the heart in sudden unexpected death due to coronary heart disease. Acta Pathol Microbiol Scand 89:157–63. 36. Thomas, A. C. and S. Pazios. 1992. The postmortem detection of coronary artery lesions using coronary arteriography. Pathology 24:5–11. 37. Weman, S. M., P. J. Karhunen, A. Penttilä, A. A. Jarvinen, and U. S. Salminen. 2000. Reperfusion injury associated with one-fourth of death after coronary artery bypass grafting. Ann Thorac Surg 70:807–12. 38. Von Wedel, J., J. W. Lord, Jr., C. G. Neumann, and N. W. Hinton. 1955. Revascularization of the heart by pedicled skin flap; an experimental study investigating the functions of the extracoronary anastomoses. Surgery 37: 32–53. 39. Böhm, E. 1982. Ergebnisse postmortaler Organ- und Gewebsperfusion. Beitr Gerichtl Med 41:449–58. 40. Weman, S. M., U. S. Salminen, A. Penttilä, A. Mannikko, and P. J. Kartunen. 1999. Post-mortem cast angiography in the diagnostics of graft complications in patients with fatal outcome following coronary artery bypass grafting (CABG). Int J Legal Med 112:107–14. 41. Grabherr, S., Gygax, E., Sollberger, B., et al. 2008. Twostep post-moretem angiography with a modified heart-lung machine: Preliminary results. AJR 190:345–351. 42. Weitzman, D. 1964. Post-mortem coronary arteriography and its correlation with electrocardiography. Br Heart J 26:330–36. 43. Hintze, A. 1931. Die Verteilung des Gefäßinhaltes beim überlebenden menschlichen Organismus und beim Versuchstier unter verschiedenen physikalischen und chemischen Bedingungen. Die Darstellung der Befunde im Röntgenbild. Virchows Arch Path Anat 281:526–700. 44. Jorgensen, S. M., O. Demirkaya, and E. I. Ritman. 1998. Three-dimensional imaging of vasculature of vasculature and parenchyma in intact rodent organs with x-ray micro-CT. Am J Physiol 275:H1103–14. 45. Djonov, V., M. Schmid, S. A. Tschanz, and P. H. Burri. 2000. Intussusceptive angiogenesis: its role in embryonic vascular network formation. Circ Res 86:286–92. 46. Yonas, H., M. Boehnke, and S. Wolfson. 1982. Radiopaque silicon rubber and xeroradiograhy for the high-resolution visualization of the cerebral vasculature. Surg Neurol 17: 130–31. 47. Li, D., R. P. Dolan, R. C. Walovitch, and R. B. Lauffer. 1998. Three-dimensional MRI of coronary arteries using an intravascular contrast agent. Magn Reson Med 39:1014–18. 48. Plachta, A., S. A. Thompson, and F. D. Speer. 1955. Pericardial and myocardial vascularization following cardiopericardiopexy. Arch Path 59:151–61.
Postmortem Angiography
451
49. Stein, B. M. and G. T. Svare. 1963. A technique of postmortem angiography for evaluating arteriosclerosis of the aortic arch and carotid and vertebral arteries. Radiology 81:252–56. 50. Wagner, A. and C. A. Poindexter. 1949. Demonstration of coronary arteries with nylon. Am Heart J 37:258–66. 51. Stern, H., E. R. Ranzenhofer, and A. A. Liebow. 1954. Preparation of vinylite casts of the coronary vessels and cardiac chambers. Lab Invest 30:337–47. 52. McGill, H. C., B. W. Brown, I. Gore, et al. 1968. Report of committee on grading lesions, council on arteriosclerosis, American Heart Association. Circulation 38:460–68. 53. Krucker, T., A. Lang, and E. P. Meyer. 2006. New polyurethane-based material for vascular corrosion casting with improved physical and imaging characteristics. Microscopy Research and Technique 69:138–47. 54. Scott, R. W., A. F. Aoung, H. A. Zimmermann, et al. 1949. An improved method for visualizing the coronary arteries at post mortem. Am Heart J 38:881–88. 55. Davis, N. A. 1963. A radioisotope dilution technique for the quantitative study of coronary artery disease post-mortem. Lab Invest 12:1198–203. 56. Cocchetti, E. and I. Donini. 1954. La senilizzazione delle arterie del testicolo nell’ uomo. Ateneo Parmese 25:318–36.
D6.2
POST-MORTEM ANGIOGRAPHY AFTER VASCULAR PERFUSION WITH DIESEL OIL AND A LIPOPHILIC CONTRAST AGENT Silke Grabherr
D6.2.1 METHODOLOGICAL CONCEPT This method of performing a whole-body angiography was developed with the following considerations: To perform an angiography that should resemble clinical in-vivo angiography, circulation is necessary to have conditions similar to in vivo. To reach this goal, the following three problems had to be solved: 1. There must be a perfusate instead of the blood. 2. A pump is necessary to replace the heart and engage the circulation. 3. A contrast agent that is suitable to the perfusate has to be injected. D6.2.1.1
Perfusate
As described in the literature, oily perfusates have the ability to remain intravascular. [1] Therefore they are suitable for performing a perfusion [2] without high loss of the perfusate into the surrounding tissue, and without edematization of it. The use of water-soluble liquids will not allow perfusation of the vascular system because they penetrate the vascular wall rapidly, a feature which is useful in embalming bodies, but causes edema [3 4]. Another ability of oily liquids is microembolization. Depending on the viscosity of the oil used, peripheral
© 2009 by Taylor & Francis Group, LLC
!
FIGURE D6.2.1 Experimental setup of the two-step method: The perfusate (center, top) consisting of diesel oil or paraffin oil, eventually colored with Sudan’s red, is transported by a roller pump (left, top). For the human corpse, a modified heart-lung machine (right, top) is applied, which allows strict pressure control. The perfusate enters the vascular system of an animal cadaver (bottom left) or a human corpse (bottom right) via a cannulated artery. After starting the perfusion, it leaves the vascular system together with remaining blood and postmortem blood clots (center, bottom) through a cannulated vein.
vessels are occluded by fatty emboli, as is known from anticancer treatment using oily liquids [5]. In our method we used diesel oil and later liquid paraffin oil because of their low viscosity. With these oils it is possible to reach vessels up to a diameter of 50 μm. Our own investigations on a chorio-allantioc-membrane (CAM) model demonstrate this mechanism of fat embolism in the microvascular system. However, the perfusate reaches the venous system by using arteriolo-venous shunts. To make the colorless oils better visible, we colored them with Sudan red. D6.2.1.2
Pump
To engage postmortem circulation, a pump is necessary. For a continuous perfusion, a roller pump is suitable. In later investigations on the human model we used a modified heartlung machine, which allows a pressure-controlled perfusion that simulates in-vivo conditions optimally.
452
The Virtopsy Approach
FIGURE D6.2.2 3D reconstructions of MSCT-data of a whole-body angiography of a dog cadaver. (a) Whole-body angiogram, obtained 2 (top) and 3 (bottom) minutes after contrast-agent application. Two minutes after contrast-agent injection, the arterial part of the vascular system is visible (white arrow: right carotid artery). The liver (red arrows) has reached its parenchymateous phase and perfusion of the kidneys has started (yellow arrows). One minute later, the parenchymatous phase of the liver is finished and the inferior portal vein (blue arrow) is visible. The kidneys (yellow arrows) are better perfused than before. By regarding the periphery vessels, like the vessels of the left ear (white dotted arrow), a further penetration of the perfusate can be recognized. (b) After virtual cutting and zooming in, the aortic arch (white arrow) is represented. In this area, the bicarotid truncus (blue arrow) has its origin, and is divided into the two carotid arteries (yellow arrows). (c) Demonstration of the kidneys, 3 minutes after contrast agent injection.
D6.2.1.3
Contrast-Agent
To be suitable to the oily perfusate, a lipo-soluble contrast agent was chosen. Lipiodol Ultrafluide®, which can be found in the literature [6], is compatible and has a high opacity of around 2000 HU.
D6.2.2 TWO-STEP METHOD Initially, a post-mortem circulation is established by preparation of the body and perfusing the vascular system with an oily fluid. The vessels are then rendered visible by injecting a lipophilic agent into the established circulation. The experimental setup is demonstrated in Figure D6.2.1. D6.2.2.1 Establishing a Postmortem Circulation Principally two vessels have to be cannulated: one artery to introduce the vascular system, and one vein through which the
© 2009 by Taylor & Francis Group, LLC
perfusate should leave it. To choose the ideal vessel it should be considered which part of the vascular system will be demonstrated, or if vessels of the whole body are of interest. In our animal studies we performed the following perfusions: Cranial circulation: Cannulation of the abdominal aorta and inferior vena cava from caudal to cranial Caudal circulation: Cannulation of the abdominal aorta and inferior vena cava from cranial to caudal Whole body circulation: Cannulation of the carotid artery and jugular vein from one side from caudal to cranial In human bodies the following perfusions were performed: Circulation of one leg: Femoral artery and femoral vein from one leg from cranial to caudal
Postmortem Angiography
Retrograde venous perfusion: To demonstrate the veins of an arm in a deceased person with suspicious death because of an intravenous overdose of drugs, the subclavian vain from one side has been cannulated.
453
Important for preparation is a careful occlusion and ligation of small side branches of the cannulated vessels to avoid leakage and loss of the perfusate. The incision on the body can be performed as minimally as possible. After cannulating and connecting the arterial cannula with the roller pump or the modified heart-lung machine, perfusion was started by filling the tube system with the colored oil and engaging the pump. Shorty after starting the perfusion, the oily liquid appears in the venous cannula and flushes out together with the remaining blood and postmortem blood clots.
membrane applicator as a bolus injection. We injected 40 ml for angiography of an animal cadaver or a human leg. One or two minutes after the bolus injection, first images should be obtained. The imaging technique can be chosen variably. In our studies, we performed conventional x-ray angiography as well as multislice computed tomography (MSCT) angiography. To reach a dynamic angiography, which shows the arterial, parenchymatous, and venous systems, images are obtained in different time intervals, in analogy to cinical in-vivo angiography. Depending on the perfused body and the perfusion pressure, the different parts of the vascular system are demonstrated after different intervals, and after rinsing through the vessels, contrast agent is flushed out of the body. Two or three captures performed within 2-minute time frames, seem to be a good choice.
D6.2.2.2 Angiography
D6.2.3 RESULTS OF THE TECHNIQUE
After the circulation has been running for about 2 or 5 minutes, angiography can be performed in a second step. Lipiodol Ultrafluide is injected into the still-running circulation. The injection can be set into the tube system with an inserted
By establishing a postmortem circulation, a dynamic angiography showing the arterial, parenchymatous and venous system can be performed in analogy to clinical in-vivo angiography. The vessels are perfused up to the level of arterioles. Figures D6.2.2 6
FIGURE D6.2.3 Demonstration of the dynamic process of the method, using 3D reconstructions of MSCT-scans. (a) MSCT angiograms of a dog, 3 days after death. The right image was obtained 4 minutes after contrast agent injection and is showing the arterial phase (yellow arrow: right carotid artery; red arrows: branches of the facial artery). The venous phase is visible on the left image, obtained 9 minutes after the injection of Lipiodol Ultrafluide (yellow arrow: left jugular vein; red arrow: facial vein). (b) MSCT angiograms of a cat, 2 days postmortem. The right image, obtained 2 minutes after contrast-agent injection, represents the arterial phase of the angiogram with small arterial branches, supplying the musculature of the hind limbs. On the left side, an angiogram obtained 9 minutes after contrast-agent injection shows the veins (red arrows: iliac veins; yellow arrows: saphenous veins), while the contrast agent is already washed out of the arterial system.
© 2009 by Taylor & Francis Group, LLC
454
The Virtopsy Approach
FIGURE D6.2.4 3D reconstructions with changed windowing demonstrate the high opacity of the contrast agent. By using a window higher than bone opacity, an overview of the great vessels can be established (white arrows: aorta; yellow arrow: bicarotid truncus; white dotted arrows: carotid arteries; yellow dotted arrows: maxillar arteries; blue arrows: brachial arteries).
FIGURE D6.2.5 MSCT angiogram of a human leg, performed 2 days postmortem on a corpse showing macroscopic signs of peripheral arterial occlusive disease. (a) By regarding the leg from behind, clear discrepancies are visible between the number of perfused vessels femoral and at the level of the shank, better visible on the right image, with softer windowing. On the level of the popliteal cave (white arrow: popliteal artery), the perfusion decreases. The posterior tibial artery shows an abrupt stop of perfusion (red arrows).
© 2009 by Taylor & Francis Group, LLC
Postmortem Angiography
455
FIGURE D6.2.5 (CONTINUED) (b) By zooming in on the popliteal space, the vessels can be examined in detail. The fibular artery (black arrow) is divided shortly after it’s origin into many small branches. The posterior tibial artery shows a complete occlusion (red arrow).
FIGURE D6.2.6 Postmortem venous MSCT angiography of the right arm of a 39-year-old male with known history of intravenous drug abuse and a visible injection mark with surrounding hematoma in the right cubital space. (a) 3D angiogram showing the skin and superficial veins of the right arm (yellow arrow: cannula placed in the right subclavial vein). (b) Maximum-intensity projection showing a longitudinal slide of the right cubital space (area inside the yellow quadrate of (a)). The medial cubital vein (red arrow) is cut longitudinally. Along the vein, extravasations of contrast agent are visible (red arrows). The skin above the vein shows a little hump of discrete swelling (white arrow). (c) 3D reconstruction of the right arm from internal and dorsal view, showing the veins of the arm (white arrow: brachial vein; yellow arrow: medial cubital vein; white dotted arrow: cephalic vein). (d) By zooming into the cubital space of the 3D model (red quadrate in (c)), an extravasate of contrast agent (black arrows) is visible above the medial cubital vein, which corresponds with the localization of the injection mark and the hematoma, described above.
© 2009 by Taylor & Francis Group, LLC
456
The Virtopsy Approach
FIGURE D6.2.7 3D reconstruction of postmortem MSCT angiography showing the vasculature of the left hand from dorsal (right) and palmar (left) views in detail. The images were obtained 3 minutes after contrast-agent injection.
show examples of results reached with this technique. The method is practicable on human bodies as well as on animal cadavers, and on body parts as well as on whole bodies. After angiography, the contrast agent is flushed out, and with a second bolus injection of contrast agent, angiography can be repeated. Due to fatty microembolisms, the microvascular system is not perfused, thus making angiography possible even several days after death. While in the literature a limit of 24 hours is generally described, our method was used successfully even 5 days after death [7] without visible edematization of the body. Time between injection of the contrast agent and imaging can be extended when the circulation is stopped after contrast agent injection, due to the intravascular resting of the oily Lipiodol ultrafluide. This fact can be useful if preparation and contrast agent injection is logistically or ethically not possible at the place of imaging.
4. Macdonald, G.J., Macgregor, D.B. Procedures for embalming cadavers for the dissecting laboratory. Proc Soc Exp Biol Med 1998; 215:363 365. 5. Nakakuma, K., Tashiro, S., Hiraoka, T., et al. Studies on anticancer treatment with an oily anticancer drug injected into the ligated feeding hepatic artery for liver cancer. Cancer 1983; 52:2193 2200. 6. Pfeifer, K. J., Klein, U., Chaussy, C. H., et al. Postmortale nierenvergrößerungsangiographie mit fettlöslichem kontrastmittel. Fortschr Röntegenstr 1974; 121:472 476. 7. Grabherr S. Postmortaler Kreislauf mit angiographischer Darstellung der arteriellen, kapillären und venösen Strombahn. Thesis, Medical University of Innsbruck, Austria; 2004.
D6.3
POSTMORTEM MINIMALLY INVASIVE ANGIOGRAPHY Christian Jackowski and Michael J. Thali
D6.2.4 REFERENCES 1. Zapata, M. G., Alcaraz, M., Luna, A. Study of postmortem blood circulation. Z Rechtsmed 1989; 103:27 32. 2. Barmeyer, J. Postmortale koronarangiographie und perfusion normaler und pathologisch veränderter Herzen, Messung der Durchflusskapazität interkoronarer Anastomosen. Beitr Pathol Anat 1968; 137:373 390. 3. Grabuschnigg, P., Rous, F. Preservation of human cadavers throughout history: A contribution to development a methodology. Beitr Gerichtl Med 1990; 48:445 458.
© 2009 by Taylor & Francis Group, LLC
As postmortem investigation was increasingly supported by computed tomography (CT) and magnetic resonance imaging (MRI) within the Virtopsy Project and further research groups [1–14], the idea was born to also implement a noninvasive or minimally invasive technique for vessel visualization. Unfortunately, cross-sectional visualization of the pathologies of the vascular system remained a challenge, and without an adequate vascular diagnostic tool the virtopsy approach could not reach the goal of serving as an
Postmortem Angiography
457
D6.3.1
EX VIVO PORCINE MODEL
Preliminary experiments to visualize arteries using multislice CT (MSCT) and MRI as well as the microvascular circulation using MRI were therefore carried out on the coronary artery system of ex vivo porcine hearts. Specimens were purchased from a local slaughterhouse. Dissection contained the preparation of the aortic root and the ligature of all further major vessels to avoid leakage of the tested contrast agents, which could have caused epicardiac artifacts. The aortic root was intubated via the brachiocephalic artery with a bulbous probe and tightly closed by several enlacements (Figure D6.3.1(A)). The prepared hearts were placed in plastic bags to avoid contamination of the radiological equipment. A native MSCT scan was then performed. Prior to injection of the contrast medium, the air within the heart that could not be avoided due to the slaughtering procedure was extracted via the probe using a conventional syringe. 100 ml to 120 ml of contrast agent (barium- or iodine-based) at different concentrations was injected. MSCT scanning was performed on a 16-row scanner (Sensation 16, Siemens) with a collimation of 16 × 0.75 mm, calculated slice thickness of 0.8 mm, and an increment of 0.3 mm. For MRI the procedure remained the same except for a gadolinium-based T1-shortening contrast agent (Dotarem) in an aqueous solution replacing the x-ray contrast agent. MR scanning was performed on a 1.5 Tesla system (Magnetom Sonata, Siemens), and sequences were as follows: axial T1-weighted (turbo spin-echo [TSE], echo time [TE]-17 ms/ repetition time [TR]-656 ms, flip angle 180°, slice thickness
Sonata Maestro Class
A
MAGNETOM
all-embracing alternative postmortem investigation technique to the traditional forensic autopsy in selected cases. Postmortem angiography has a long history, and several approaches of postmortem visualization of pathologies within the vascular system have been described. Twodimensional x-ray projection angiograms using barium sulfate solutions [15–18], gelatine barium suspensions (19-24), silicone rubber charged with lead oxide [25–36], gastrografin [37], or iodinated contrast agents [38,39] were used to answer specific questions regarding the vascular system. In 2001, Rah et al. [40] presented the first postmortem 3D coronary angiogram using electron-beam computed tomography. Common ground of the entire literature concerning postmortem angiography is that it was based on invasive preparation techniques and selective intubation of the vascular system to be examined, and therefore depended on a traditional autopsy. The aim of the Virtopsy Project in Bern [1,41] was to implement a minimally invasive autopsy technique for postmortem investigations. In contrast to the previously published literature we started first with experiments investigating today’s feasibility of different contrast agents and contrastagent solutions for postmortem CT examinations. Most of them either already have been used for invasive postmortem x-ray examinations in the past or are routinely used in clinical radiology. As the human heart only differs in dispensable details from the porcine heart, porcine hearts were chosen as the model to address such initial questions as what contrast agent to use, how to use it, what concentrations to use, and how to inject [42].
B
FIGURE D6.3.1 Porcine heart model: (A) Preparation of the porcine hearts with a plastic bulbous probe in the supravalvular ascending aorta and a three-way stopcock. Further vessels have been ligated (arrows) or tightly sutured. (B) For MR examination the porcine heart was placed in a conventional knee coil (usually used for MR examinations of the extremities). Injection of the gadolinium solution was performed via (A) the tube connected the three-way stopcock. For CT examinations the setting remained comparable except for the coil.
© 2009 by Taylor & Francis Group, LLC
458
The Virtopsy Approach
A
B
f e a
d
b c
g
FIGURE D6.3.2 Postmortem coronary MSCT angiography using the ex vivo porcine model: (A) The four images show an oblique anteriorposterior view on a volume-rendered (VR) 3D model with successive decreasing of the soft tissue’s opacity. Note also that the ascending aorta becomes displayed as the injection of the meglumine ioxithalamate was performed into the aortic root. (B) Cranio-caudal view of the reconstructed coronaries. (a) Left anterior descending (LAD) coronary artery. (b) First diagonal branch. (c) First septal perforator. (d) Circumflex (CX) coronary artery. (e) First posterolateral branch. (f) Second posterolateral branch. (g) Right coronary artery.
intubation of the coronary orifices (Figure D6.3.2). The latter is especially important, as it is not expected that it is possible to intubate the coronary orifices at an intact corpse in a minimally invasive way with reasonable efforts. These results showed that this is not necessary at all. Concentrations above 10% were useful for MSCT. MRI results were inferior to MSCT concerning the visualization of the coronary system because of the increased slice thickness that was applied, but nevertheless a sufficient display of the three main coronary arteries in adequate quality using a 1% aqueous solution of the administered gadolinium was reached (Figure D6.3.3(A)). Selective intubation of the left main coronary artery as mentioned already visualized the perfused myocardium in an
3 mm) and fluid-attenuated inversion recovery (FLAIR) 3D (TE-1.4 ms/TR-3.8 ms, flip angle 25°, slice thickness 1 mm) for high-resolution images. To visualize the intramyocardial microcirculation by MRI, the preparation differed as follows: In a series of two hearts the aortic root was dissected, and the coronary orifice of the left coronary artery was intubated with a smaller bulbous probe and tightly ligated. Injection of the gadolinium at a concentration of 0.5% was thereby limited to the myocardium fed by the left main coronary artery. As a result of the ex vivo porcine model, MSCT and meglumine ioxithalamate (Telebrix) provided an excellent visualization of the coronary artery system, without selective
A
B c b
Anterior
a
d Posterior FIGURE D6.3.3 Postmortem coronary MR angiography using the ex vivo porcine model: (A) Cranio-caudal view on a 3D VR reconstruction of the porcine coronary artery system injected with a gadolinium solution via the aortic root. (a) Left anterior descending coronary artery. (b) Circumflex coronary artery. (c) First posterolateral branch. (d) Right coronary artery. Specimen contained small amounts of air due to the slaughtering procedure that caused contrast-agent defects within the vessel and thereby simulate occlusion of the right coronary artery as an artifact (*). (B) Short-axis T1-weighted image (TE-15 ms/TR-400 ms) after selective injection of the left main coronary artery. The myocardial distribution of the contrast agent depending on the anatomy of the left main coronary artery is clearly depicted.
© 2009 by Taylor & Francis Group, LLC
Postmortem Angiography
A
459
B
C
FIGURE D6.3.4 Human model: (A) Angiography of the whole human corpse was performed via a small right inguinal incision to get access to the right common femoral artery. A plastic tube was advanced into the aortic arch in guide wire technique. (B) Using a T-piece within the left common femoral artery a pressure control during injection could be performed without cutting off the left leg from the angiography. Preparation was carried out in the autopsy room, and when finished the corpse was wrapped as usual for cross-sectional examinations using the equipment of the radiological department. The tubes were allowed to leave the wrapping. (C) Setting of the whole corpse wrapped twice for angiography using MSCT. The flexible tubes leaving the wrapping were connected to a flow-adjustable pump and a conventional manometer. During injection via the right femoral artery into the ascending aorta the arterial pressure could be monitored using the manometer connected to the left femoral artery. That allowed limiting the intravascular pressure to 60 mmHG during injection.
excellent contrast to the remaining myocardium on axial T1-weighted images (Figure D6.3.3(B)). This will become important in the future when coronary narrowing or occlusions need to be assessed concerning their relevance for the local myocardial microcirculation.
D6.3.2 HUMAN CORPSE MODEL Knowing the agent to use and the most useful concentration, three nonfixed and nonforensic corpses were initially investigated using MSCT in close collaboration with the Institute of Human Anatomy at the University of Bern. To visualize the human arterial system including the coronary arteries in a minimally invasive way, both common femoral arteries were dissected in supine position (Figure D6.3.4(A)). Using the Seldinger technique, a rigid guide wire was advanced into the aortic arch via the right femoral artery. Using the guide wire, a flexible tube of the maximal possible, diameter, ranging from 5 mm to 8 mm, was also advanced up to the aortic arch and tightly closed with a clamp. The tube was fixed to the femoral artery by several ligating enlacements. Two of the three human cases showed severe atherosclerosis of the iliac arteries, complicating the placing of the tube. In these cases the problem was solved by simultaneous use of an intravascular catheter and a more flexible guide wire. After having placed the catheter correctly, the flexible guide wire was replaced by a back-up guide wire of distinctively lower flexibility. The catheter was then removed, and the
© 2009 by Taylor & Francis Group, LLC
backup guide wire was used to advance the flexible tube into the aortic arch. A T-piece was inserted into the left femoral artery (Figure D6.3.4(B)) and tightly fixed to monitor the intravascular pressure while the injection of the contrast agent was performed. A conventional manometer was connected to the third access of the T-piece after the air within the system was removed. The preparation was performed in the autopsy room, and afterward the corpse was wrapped into two body bags that had been proven to cause no imaging artifacts. The tubes left the body bags via two small incisions. In the MSCT scanning room, the injection tube was connected to the tube, leaving the flow-adjustable injection pump (COBE perfusion system, Lakewood, CO) as used also in heart-lung machines. The injection system was completely filled with contrast medium to ensure that no air remained within the system or could enter it. Postmortem whole-body angiography was performed using an aqueous solution of meglumine ioxithalamate at a concentration of 20%. The radiological examination started with a native scan. Thereby arterial calcifications could be detected without overlaying contrast within the vessel lumen. The contrast agent was then injected on the scanner table (Figure D6.3.4(C)). We started with 15–20 ml/kg body weight. The injection flow should be adapted to vital conditions of a regular cardiac output but limited by an intravascular pressure not to exceed 50–60 mmHg. The flow was therefore slowly increased to a maximum of
460
The Virtopsy Approach
A
B
C
D
FIGURE D6.3.5 Whole-corpse CT angiography: (A) Photographic documentation compared with the VR angiographic CT data (B–D) of the case. (B–D) 3D VR models with successive decrease of soft-tissue opacity to display the contrast-agent-filled vascular system. Note that the contrast-agent-filled tube in the right inguinal vascular access is displayed in all four images (arrow). In contrast, the left inguinal access is not visible within the VR CT data as it solely contained air that was set to 0% opacity (dashed arrows). However, in (C) and (D) the left femoral artery becomes visible as is also contrast-agent filled (doubled arrows). Furthermore, a beginning venous overlap can be noted in the jugular veins as a result from an inadequate large contrast-agent volume (arrow heads).
3–4 l/min, resulting in an injection time of 40–60 sec. Scanning was started immediately after the end of injection before a collapse of the vessels could occur. A second or even a third injection with subsequent scanning was performed when gas bubbles occurred within the vascular system in order to displace the gas and to differentiate these as artifacts. The time needed for preparation of the corpses, logistics, injection, and scanning ranged from 2 to 3 hours. After angiographic investigation, the corpses were partly dissected to validate the postmortem angiographic findings.
A
Postmortem angiography of the three human corpses was successful in visualizing the entire human arterial system, including the intracranial arteries (Figure D6.3.5). A complete visualization of the arterial circle of Willis as well as its branches is exemplarily shown in Figure D6.3.6 and is displayed in more detail concerning the peripheral branches than routine autopsies can provide. Three-dimensional reconstructed models allow for a fast overview of the vascular system, and detection of relevant narrowing as seen in Figure D6.3.7 on a stenosis of the left carotid artery becomes easier. In addition to the detection of stenosis, its density and composition
B h g f e d c
FIGURE D6.3.6 Cranial 3D VR MSCT angiography (same case as in Figure D6.3.9): (A) Posterior-anterior view visualizes the intracranial arterial system on a slab from the whole cranial data set; slab thickness and orientation is indicated in (B) (dashed frame). (B) Lateral view on a cranial slab of the same data set as in (A); slab thickness and orientation is indicated in (A) (dashed frame). Note the vertebral arteries (c), the basilar artery (d), a small posterior cerebral artery (e), the arterial circle of Willis (f), anterior cerebral artery-pericallosal artery (h), and middle cerebral artery (g).
© 2009 by Taylor & Francis Group, LLC
Postmortem Angiography
461
A
D
C
B
E
F
G
FIGURE D6.3.7 Stenosis of the left internal carotid artery (same case as in Figure D6.3.4, Figure D6.3.5, and Figure D6.3.8): (A) Native axial MSCT image shows two calcifications within the wall of the left internal carotid artery (arrow). Right common carotid artery also shows calcifications (dashed arrow) that appear even more severe compared with the left side. (B) However, postcontrast axial MSCT image visualizes the lumen of the left internal carotid artery distinctively narrowed between both calcifications (arrow) compared with the left external carotid artery (*). Angiography of the right common carotid artery reveals a normal lumen (dashed arrow) between the calcifications. Note also that the contrast agent already entered the venous system. (C) 3D VR reconstruction of the left carotid artery illustrates the narrowing just above the bifurcation. (D) Magnification of the left internal carotid artery (dashed circle) in (A). (E) Magnification of the left internal carotid artery (dashed circle) in (B). (F) Histological cross-section (hematoxylin and eosin, or H&E) of the presented stenosis. Calcifications (blue arrows) appear dense and therefore white within the vessel wall at MSCT (D, E). The lipid rich core of the lesions (yellow arrows) appears dark at MSCT (E) as the density of fat (~–100 Hounsfield units, or HU) is distinctively lower than that of other soft tissues (0–100 HU). The narrowed lumen (black dashed arrow) is bright at MSCT due to the radioopaque contrast agent. (G) Autoptical appearance of the stenosis is shown (arrow).
could be correlated to the histological appearance and showed hyperdense calcification parts and hypodense fatty areas (Figure D6.3.7(D–F). For more detailed explanation, see the figure legends. Minimally invasive coronary angiography also became possible as expected after the porcine experiments that succeeded well without intubation of the coronary orifices (Figure D6.3.8). The main coronary branches became visualized and allowed for calcified plaque detection and for an assessment of the vessels’ patency. On the other hand, softtissue injury could also be assessed by extravasation of contrast medium, as seen in a case with an agonal bruise on the left forehead (Figure D6.3.9).
D6.3.3 THE METHOD AND THE EXISTING LITERATURE Postmortem angiography has a long history and goes back to the early decades of the 20th century [43–48]. The existing literature is based on traditional autopsies as it
© 2009 by Taylor & Francis Group, LLC
was used to investigate the vascular systems of isolated organs, such as the heart [16–23,31,37,38,43,46,48–103], the brain [29,30,32,34,36,104–115], the lung [116,117], the kidneys [47], the spleen [33], the intestine [27,118–120], the uterus [121], the spinal column [122], and the extremities [123,124]. Exceptions to this are postmortem angiographic investigations of fetuses and newborn babies [39,45,125– 131], as the umbilical vessels were predominantly utilized for injection of the radioopaque contrast agent without previous autopsy. Impressively, these studies imaged the entire arterial system of the fetus, and this triggered once more the idea to transfer this minimally invasive angiographic approach onto adult human corpses in order to assess the vascular pathology within the concept of a minimally invasive autopsy. Contrary to fetuses and newborn babies, adult human corpses do not present with an existing vascular access. Therefore, an access to the arterial system had to be created, and we used two small inguinal incisions for that purpose. We
462
The Virtopsy Approach
A
B
C
E
F
G
D
FIGURE D6.3.8 Minimally invasive postmortem coronary MSCT angiography presented as a comparison of 3D VR data (left), autopsy specimen (middle), and curved reformatted CT data (right): (A–D) Left main coronary artery branching into a left anterior descending, a major intermediate branch, and a small circumflexa (arrow with **). Calcifications of the left orifice and within the proximal LAD cause narrowing of the vessel (dashed lines). Nevertheless, as the calcifications nowhere totally occlude the LAD its distal segments are well visualized and show no further pathology. (E–G) Right coronary artery; note additional calcifications within the course of the right coronary artery (dashed lines). The (*) indicates small gas bubbles causing artifacts within the data.
deemed this as acceptable, as similar incisions are performed in embalming or conservation procedures that also aim at keeping the body intact [132–134]. Future alternatives might be less invasive by using semiautomatic image-guided puncture systems, which have not yet been implemented, to reach the aortic arch for injection of the contrast agents [135–138]. When these methodical challenges are overcome, the open preparation procedure, as presented here, will become obsolete. Contrary to our own expectations, it was in all of the three cases possible to get the flexible tube via the femoral and iliac artery through the abdominal aorta into the aortic arch. In two of the three corpses the procedure was slightly complicated by severe atherosclerosis of the iliac artery, but the technique as described could easily overcome this problem. The access via the femoral artery has the disadvantage that it cuts off the right femoral artery from the angiography, as a tight ligation is needed to fix the tube within the vessel. Thereby the right leg was excluded from being injected. This current problem should also be overcome when the technical conditions for an automated transthoracic puncture are implemented. A possible method might be a puncture of the ascending aorta under image guidance through upper intercostal spaces, thus simplifying the method. In the second vascular access for the pressure control, a T-piece with two ligations was inserted, thus allowing the contrast medium to
© 2009 by Taylor & Francis Group, LLC
reach the periphery of the second leg (Figure D6.2.4(b)). This might also become obsolete when a puncture for the placing of a pressure measurement probe becomes possible. Discussing the minimal invasiveness of the approach, we questioned whether the visualization of the coronary arteries is sufficient for diagnostic purposes without separate intubation and injection of contrast medium in each coronary orifice. As long as an intraaortic pressure that reliably presses the contrast medium through the coronaries can be achieved, a selective orifice intubation seems not to be necessary. In the literature, several studies have shown adequate results in coronary and bypass graft visualization by injecting the contrast medium into the aortic root [62,64,94]. Selective intubation of the coronaries may result in an increased image quality but will no longer be practicable in a minimally invasive manner. This problem will be nearly intractable when aorto-coronary bypass grafts need to be investigated. Injected into the ascending aorta with a low physiological arterial pressure of 50–60 mmHg, the contrast medium will distribute within the entire arterial system including prior unknown bypass grafts.
D6.3.4 THE INJECTION PRESSURE In the literature, various injection pressures are advised. Contrast media of increased viscosity require an elevated
Postmortem Angiography
A
C
463
B
D
FIGURE D6.3.9 Cranial postmortem CT angiogram in a case with an agonal crush wound (same case as in Figure D6.3.6): (A) Note the crush wound present on the left forehead (arrow). (B) Anterior-posterior view on the VR CT data shows distinctive extravasation of contrast agent within the tissue above the left orbital region. (C) Left lateral view. (D) Cranio-caudal view. Note that the contrast-agent volume was distinctively better adjusted to the need for the case, as no venous overlapping is visible within the data (compare with Figure D6.3.5 and Figure D6.3.7).
injection pressure compared with vital conditions, ranging from 150 mmHg [21,23,51] to 200 mmHg [57,61,68,118] and up to 250 mmHg [47]. Aqueous contrast agents of lower viscosity need lower injection pressures such as 120 mmHg [17,109], 100 mmHg [75] or less [139]. It is not advisable to reach these high pressures when angiography of an entire corpse is performed. The capillary system of the intestinal wall in particular, being first exposed to the destructive putrefaction process, might not withstand “vital” pressures. According to our experience, even at pressures of 60–70 mmHg the contrast medium can enter the bowel wall and penetrate into the intestinal lumen or cause a diffuse enhancement of the autolytic pancreas (Figure D6.3.10). If the injection pressure is further increased, a relevant volume of contrast medium might get lost within the intestine and cause different artifacts instead of enhancing the vessels. Therefore, the maximal intravascular pressure during injection should be progressively reduced with increasing postmortem interval. In other words, angiographic investigations should be performed as early as possible after death.
© 2009 by Taylor & Francis Group, LLC
A
B
FIGURE D6.3.10 Special postmortem MSCT angiographic characteristics and problems to be aware of: (A) Massive enhancement of the pancreas due to autolytic vulnerability of the pancreatic capillary bed (arrow) and enhancement of the bowel wall due to beginning putrefaction (dashed arrow) is present. Note the enhancement of the renal cortex due to increased vascularization compared with the medulla. (B) Ruptures within the vulnerable capillary bed of the gastric and bowel wall cause the contrast agent to enter the gastric or bowel lumen (arrow). Note the enhancement of an intramuscular hematoma within the left latissimus dorsi muscle (*).
464
D6.3.5 THE INJECTED CONTRAST-AGENT VOLUME The volume to be injected is larger than for single organ or fetal angiography. To fill the entire arterial system to the periphery, the injected volume needs to be more than the expected arterial volume of the corpse. The contrast agent might also enter the pulmonary veins due to regurgitation via insufficient aortic and mitral valves. Thereby, further volume may get lost. On the other hand, to inject volumes distinctively larger than the arterial volume results in a venous overlap that will complicate image interpretation and should therefore be avoided. In our experience it is advisable to start with 1 to 2 liters, adapted to the sex and habitus of the corpse, and eventually increase the volume through an additional injection.
D6.3.6 THE CONTRAST AGENTS Various radioopaque agents have been utilized in the past for postmortem angiography. Barium sulfate suspended in H2O has been very popular [16–18,38,73–78,82,84–86, 105,107–110,116,121,124,126–128,130,139–142]. The authors showed high-quality angiographic x-ray images of the investigated vessels. As barium sulfate is available in different particle sizes, not all suspensions reach the capillary bed, and this may lead to variable visualization of the capillary system. As meglumine ioxithalamate, which has been used in our approach, is a water-soluble iodinated contrast agent [114,115], radiographic visualization of the capillary bed depends only on the sufficiency of the vascularization and the injection parameters. Nevertheless, we consider barium sulfate suspensions of adequate particle size an applicable alternative, because our preliminary experiments on the porcine hearts included aqueous barium sulfate suspensions and were able to show comparable results to those of iodinated components. Nearly as often, barium sulfate has been used in suspension with gelatin [19–23,47,49,51–59,62–68,71,72,81,91,102104,117,118,120,125,131,143–147]. This combination is suited for radiological investigation of extracted organs or organ parts, as it is a thin fluid when warmed up that can easily be injected into any kind of tubular system, and hardens when cooling down. Thereby the barium sulfate remains within the vessel system and can be used for angiographic or microangiographic x-ray investigations even when the organ is further dissected or cut into thin slices. The usefulness of gelatin suspensions for minimally invasive investigation of the entire human corpse is further reduced by the limitation to one injection. When gelatin has hardened, the vessels are filled with a rigid medium and do not allow for an additional injection. Additional injections are useful when small gas bubbles, whether by gas embolism or due to putrefaction, which cannot be securely excluded within the corpse, occlude parts of the vascular system. Further injections move the bubbles within the vessel to the periphery, and this can allow visualization of the lumen of a vessel previously occluded by gas. Silicone as carrier substance demonstrates similar properties to gelatin [25–36,92,94,122]. Once hardened, it can display the anatomy and the lumen of the investigated vessel
© 2009 by Taylor & Francis Group, LLC
The Virtopsy Approach
after autopsy when extracted from the vessel as a cast. For postmortem minimally invasive angiography this advantage is irrelevant, because a dissection of the vessel is not intended. Therefore, the limitation to one injection limits its use for the virtual approach. Besides hydrophilic contrast media and aqueous suspensions, lipophilic contrast media or oily suspensions have been used for different reasons to visualize the vascular system on postmortem x-rays [43,69,70,90,95–97,119,123]. These media leave the vascular system within the capillary bed to a distinctively lesser extent. Comparison of both types of contrast agents has shown that lipophilic contrast media systematically visualized larger diameters of investigated vessels compared with aqueous media [148]. Additionally, the investigated vessel diameter showed an increased standard deviation when lipophilic media were used [148]. This may be a result of increased injection pressure caused by an increased viscosity as compared with the aqueous solutions, and higher pressure may inflate the postmortem vessel diameter more during and after injection. But a further reason seems to be more likely causative for a systematic overestimation of the diameter. Lipophilic agents can easily interact with the vessel wall. Especially in vessel regions with atheromatosis and stenosis, these media easily enter the vessel wall and dilute with lipids deposed within the intima such as cholesterol [149]. Thereby extravasations of the contrast medium into the plaque can occur and cause diagnostic problems in assessment vessel particularly in those regions where angiography needs to be a definite diagnostic tool to gain routine application. Furthermore, lipophilic media show an increased temperature dependence of their viscosity, causing alternating injection pressures in contrast to the constantly low viscosity of the hydrophilic media within the range of usual corpse temperatures (0°C–40°C) [149]. Aqueous media have the disadvantage of partly leaving the vascular capillary bed system and entering the interstitial tissue. To assess vascular morphology, the CT scan has to be performed immediately after injection has stopped as long as the applied pressure results in a “vital” shape of the arterial vessels, avoiding collapsed vessels in dorsal regions of the corpse. Interstitial accumulation of the contrast media can result in histological signs of edema. As long as imaging is used in a combination of MRI and MSCT, the differentiation between real edema and angiographic artifacts can be made by assessing the native predominantly T2-weighted MRI scans for preexisting edema. Second, a comparison of the native MSCT scan with MSCT angiography will reveal tissue around vessels that accumulates contrast medium. In particular, the ability of aqueous agents to pass the capillary bed offers the possibility to assess vascular territories with different tissue distribution, depending on the sufficiency of the corresponding arterial vessel. As shown in the ex vivo porcine model, contrast distribution within the myocardium reveals the coronary anatomy. We hypothesize that local areas of missing myocardial contrast might reflect occluded coronaries in postmortem minimally invasive angiography. Otherwise, as this is might be a relevant problem in particular
465
D6.3.7 BUT HOW TO OVERCOME THE EXTRAVASATION OF CONTRAST AGENT? As the contrast agent left the circulation within the capillary bed it would be reasonable to generally prevent the contrast agent from entering the capillary bed to prevent extravisation. The idea behind the following study was based on the knowledge that the viscosity of the injected agent, the injection pressure, and the flow resistance within the vascular system are in a close connection (Hagen-Poiseuille law). Changing the viscosity under stable injection pressure conditions was expected to influence the contrast-agent distribution, especially within the vessels of smaller diameter such as within the capillary bed. So a feasible possibility of changing the viscosity of the injected agent was needed to address this problem. Polyethylene glycol (PEG) has an increased intrinsic viscosity depending on the length of its PEG chains; in addition, it is chemically inert and nontoxic. On the other hand, it is of an amphiphilic nature and thereby can also dilute the aqueous contrast agent. We wondered whether it is possible to increase the viscosity using PEG successively until the resulting flow resistance may avoid entering of the contrast medium into the capillary bed under the advised intravascular pressure conditions of no more than 60 mmHg [42]. We hypothesized that it might be possible to define a contrastmedium viscosity that ensures a selective visualization of the arterial bed up to the precapillary arterioles but that on the other hand prevents filling of the capillary bed and consecutively prevents possible tissue edema artifact [150]. Therefore, we went back to the ex vivo porcine heart model [42] to investigate different contrast-agent viscosities and their distribution within the porcine myocardium at stable injection conditions. With PEG-200 (Schärer und Schläpfer AG, Rothrist, Switzerland) at concentrations ranging from 0% to 87%, we produced contrast-agent solutions with viscosities ranging from 1.1 to 41.0 mPas (at 20°C room temperature); these are displayed in Table D6.3.1. The viscosity of each solution was analyzed using a conventional
TABLE D6.3.1 Investigated Contrast-Agent Solutions and Their Composition
© 2009 by Taylor & Francis Group, LLC
% PEG
cases, we intended to address this problem in a further experimental study.
100,0 90,0 80,0 70,0 60,0 50,0 40,0 30,0 20,0 10,0 0,0
45,0 40,0 35,0 30,0 25,0 20,0 15,0 10,0 5,0 0,0
Concentration Viscosity 20°C Viscosity 25°C
1
2
3
4 5 6 Solution 1–9
7
8
Viscosity in mPas
Postmortem Angiography
9
FIGURE D6.3.11 PEG concentrations and graphical delineation of the resulting viscosities (at 20°C and at 25°C). At higher temperatures the content of PEG needs to be increased to reach the same viscosity.
viscosimeter (Viscotester VT 01, Haake, Germany) at 20°C and at 25°C, given in Figure D6.3.11. Haemalum (Mayer’s haemalum solution, Merck KGaA, Germany) was added to the contrast-medium solution to determine in histological investigations the vessels that contain contrast medium. To reach preferably high PEG concentrations within the contrast-agent solution, we limited the hemalum to a comparably low 4.3%. Under defined pressure conditions—namely, not exceeding 60 mmHG—the contrast-medium solutions were injected into the aortic root of the porcine hearts. With manual adaptation of contrast-agent flow, the pressure of 60 mmHg was maintained for 30 seconds. After injection, a high-resolution MSCT scan of each specimen was performed (Emotion 6, Siemens Medical, Germany, collimation 6 × 0.5 mm, slice thickness 0.625 mm, increment of 0.3 mm, matrix 512). As shown in the initial porcine model experiments, MSCT and meglumine ioxithalamate provided an excellent visualization of the coronary artery system. A series of shortaxis CT images of each the injected specimens with identical window settings (center 200 HU, width 400 HU) could then be compared to assess the contrast-agent distribution (Figure D6.3.12). There was no obvious change in myocardial enhancement using the lower viscosity solutions 1 to 5. From solution 6 to 9, the visualization of the myocardium was distinctively lower, with a further slight decrease in increasing PEG concentration and consecutive increase in contrastmedium viscosity. Figure D6.3.13 shows a representative histological specimen of injected specimen 1 (1.1 mPas), 5 (6.0 mPas), and 9 (41.0 mPas). In solutions 1–4, the smallest measured contrast-agent-filled vessel diameters were 7–8 μm. In solution 5 the smallest found diameter was 14.5 μm. From solution 6 onward, the filled vessels were of at least 20–25 μm in diameter. The intrinsic viscosity of vital blood is 4–5 mPas [151], and the major part of this viscosity is caused by the cellular components, as the normal viscosity of the plasma itself is somewhere around 1.3 mPas [151]. The investigated contrastmedia solutions did not contain particles with sizes comparable to erythrocytes. Therefore, the intrinsic viscosity of its
466
The Virtopsy Approach
1
2
3
4
5
6
7
8
9
FIGURE D6.3.12 High-resolution short-axis images of the porcine hearts injected with solutions 1–9 (window settings: C 200HU, W 400HU). The enhancement of the left ventricle myocardium does not obviously decrease until solution 5. From solution 6 and 7 the systematic visualization of the myocardium is prevented and the distribution of the contrast agent is limited to the coronary arteries. Note filling defects (arrows) due to air within the coronaries after slaughtering and consecutive unenhanced myocardial regions (2,5).
Length = 8.19 μm Length = 14.54 μm
Length = 27.32 μm
Length = 9.08 μm 1
1
5
5
9
9
FIGURE D6.3.13 Unstained (upper) and H&E-stained (lower) histological specimen with injection of solution 1 (1.1 mPas), solution 5 (6.0 mPas), and solution 9 (41.0 mPas). As the viscosity of the contrast solution rises, so does the smallest diameter of blood vessels into which the contrast agent was able to enter (short blue lines represent marked and measured vessel diameters). From 7–8 μm in solution 1, the minimal stained vessel diameter slightly increases to 14.5 μm; in solution 5 and in solution 9 only the terminal arterioles of 20–25 μm showed intravascular haemalum. The H&E-stained specimens are representative histological appearances of the injected porcine myocardium in solutions 1, 5, and 9. The obvious wide interstitial spaces between the myocardial fibers in solution 1 represent angiography-induced tissue edema. This phenomenon is distinctively less obvious in solution 5 and completely absent in solution 9 when the contrast-agent solution is not able to enter the capillary bed. Note the polynuclear nature of porcine myocardial tissue (arrows).
© 2009 by Taylor & Francis Group, LLC
Postmortem Angiography
solution had to be increased with the use of PEG that has an intrinsic viscosity of 40–50 mPas. The results have shown that the viscosity of injected media should be increased up to at least threefold compared with vital blood levels to prevent filling of the capillary system of the left ventricle myocardium at defined injection pressure conditions stable at 60 mmHg. Thereby the angiography-induced tissue edema seems to be reducible with increasing contrast-medium viscosity. However, further studies on human corpses are necessary to define an adequate contrast-medium viscosity under consideration of all aspects that influence the postmortem minimally invasive angiography. First of all, we expected local differences of the flow resistance in different soft tissues. This was supported by the observation in the porcine heart study that visualization of the right ventricular myocardium was not obviously influenced by increasing viscosity within the investigated ranges (Figure D6.3.12). The viscosity might be further increased to be valuable for use in a minimally invasive approach to prevent tissue edema in all soft tissues. PEGs with longer chains (e.g., PEG-400) may work comparably in larger viscosity ranges as an alternative. Otherwise, it will be important to find an adequate compromise in increasing the contrast-medium viscosity to prevent the edema artifact while still being able to detect small soft-tissue injuries such as ruptures of the liver. We expected its application on human corpses to be further complicated due to the mixture of the contrast medium with the blood of the corpse during injection, which will also influence the viscosity and thereby the behavior of the solution within the corpse. The corpse must still be turned several times prior to injection to redissolve the sedimented erythrocytes to prevent distribution of the contrast medium only within the upper serum layer. However, in cases of severe blood loss the arteries in the lacerated regions are expected to be almost devoid of blood, which of course facilitates the contrast-agent filling. PEG was used to influence the flow characteristic of this hydrophilic contrast-agent solution. Since PEGs are also used as solubilizers, they might have the ability to influence the intramural lipids of the vessel wall in atheromatosis. In systematic future experiments we need to gain more knowledge about possible PEG–vessel wall interactions, and a study may have to be undertaken to explore this. It would be inappropriate to use if it turned out to have the same disadvantages in visualization of fatty vessel wall alterations that lipophilic contrast agents have. Our investigations were performed at a room temperature of 20°C. As seen in Table D6.3.1 and Figure D6.3.11, the higher viscosities in particular are strongly temperature dependent, with a decrease of viscosity at a higher temperature. As a result, the PEG percentage of the solution needs to be adapted to the local temperature conditions of the scanning room and the corpse. The major disadvantage of the porcine heart model was that the air within the coronaries after slaughtering could not be totally drawn off, and that caused local contrast-agent filling defects within the coronaries and consecutively within
© 2009 by Taylor & Francis Group, LLC
467
unenhanced areas of the myocardium (Figure D6.3.12). However, recapitulating these preliminary experiments, it seems reasonable to influence the contrast-agent distribution by increasing its viscosity using PEG. At least in the left ventricular porcine myocardium, 15 mPas prevents a general capillary distribution of the contrast agent at 60 mmHg injection pressure.
D6.3.8 HIGH-VISCOSITY ANGIOGRAPHY ON A HUMAN CORPSE The next consecutive step was to test its feasibility and performance on a human corpse. Once again, a human corpse from the Institute of Human Anatomy was used to perform a whole-corpse CT angiography using the high-viscosity contrast-agent solution that turned out to be feasible after the porcine heart experiments. The entire preparation and injection technique remained the same as described in Figure D6.3.4. According to the results from the porcine heart study, we decided to use a PEG concentration of 65%, which resulted in an approximated viscosity of 18 mPas, as we expected a slight mixture of the contrast-agent solution with the blood of the corpse. The injection process greatly benefited from the increased viscosity, as it was distinctively easier to maintain the 60 mmHg injection pressure compared with use of the initially published low-viscosity contrast-agent solution [42]. The minimally invasive whole-corpse angiography with PEG as contrast-agent dissolver provided an excellent demonstration of the arterial system of the corpse in 3D that distinctively exceeded the angiographic quality of the initial study (Figure D6.3.14). Also, the intracranial vessels (Figure D6.3.15) and the coronaries (Figure D6.3.16) were displayed as far as the smallest branches, such as septal and posterolateral branches of the coronaries. None of the artifacts due to vulnerable vessels as described within the initial study occurred, such as an enhancement of the pancreas or the contrast agent entering the bowel lumen. There was no longer a general distribution of contrast agent within the capillary bed of the body tissues as a result of the increased viscosity. However, three exceptions could be recognized. The renal cortex, the left ventricular myocardium, and the brain became enhanced (Figure D6.3.17). For the brain, the enhancement distinctively improved its visualization using CT. The injected contrast agent enhanced specifically the white matter of the brain, clearly depicting the border between white and gray matter and thereby giving an accurate anatomical visualization of the brain structures in CT. An assessment of the anatomical structures of the brain and cerebellum was provided that has so far only been reached with the use of MRI. Since MRI is much more elaborate in use and usually requires more efforts for a forensic institution to gain access to it, one should be aware of this possibility to increase brain diagnostics in CT. We appraised a centralization of the circulation during the dying process with maximal dilatation of the precapillary arterioles in the brain, the renal cortex, and the left
468
The Virtopsy Approach
A
B
CX LAD
RCA
LAD
TDA TDA
LAD
ITA
ICA
FIGURE D6.3.16 Magnification and removal of the surrounding tissues in the thoracic data from the case seen in Figure D6.3.14 reveal normal coronary artery anatomy (LAD, left anterior descending coronary artery; CX, circumflex coronary artery; RCA, right coronary artery) in two oblique views (A,B). Neither relevant narrowing nor stenoses are present, but an obvious screw-like epicardial course of the LAD and of the posterolateral branches of the circumflex coronary artery is demonstrated, strongly indicating a chronic hypertonia. Note also the port system seen in Figure D6.3.14 correctly located within the right atrium (yellow arrow).
F
FIGURE D6.3.14 3D visualization of the arterial system in a corpse by CT and a high-viscosity contrast-agent injection. The figure shows an AP view focused on the head and trunk out of the whole-corpse angiography. Normal human arterial anatomy is demonstrated including also small branches such as the thoracodorsal arteries (TDA), the intercostal arteries (ICA), the internal thoracic arteries (ITA), or the coronary arteries (LAD, left anterior descending coronary artery). Note also a central venous port system (yellow arrows).
ventricular myocardium at the time of death as the most likely causation for this phenomenon. The postmortem angiography can thereby give a visual impression in 3D of the centralization effects on the local blood supply of these organs. As the enhancement of the brain was strongly specific within A
the white matter, the slightly amphiphilic nature of the PEG might also have contributed to it. We stated in the experimental porcine heart study [150] that it might be necessary to define an adequate contrastmedium viscosity for the corpse under consideration of all human soft tissues, as we expected local differences of flow resistance in different soft tissues. The results of the test on the human corpse partially support this. On the other hand, the used viscosity worked very well, and the local enhancement of the three organs seemed not to be a technique-related artifact but rather displayed a local change in blood circulation at the time of death. Therefore, we may conclude that the viscosity of 18 mPas used for the previously described experiment could be used for further feasibility research in larger study groups (at 20°C room temperature).
B
D6.3.9 GENERAL THOUGHTS AND PERSPECTIVES
FIGURE D6.3.15 Magnification and removal of the skull bone in the cranial data from the case seen in Figure D6.3.14 provides insight for the intracranial arteries: (A) Slight oblique AP view. (B) Right lateral view. Note the excellent representation of the entire cerebral arterial system until the most peripheral branches.
© 2009 by Taylor & Francis Group, LLC
The presented methods of minimally invasive postmortem angiography not only detect stenosis or plaques but also can detect injuries, showing external or internal extravasation of the contrast agent. Elaborate preparations to find the bleeding source might be supported by a short radiological investigation in the future. This application might also gain special importance in the visualization of soft-tissue injuries such as small liver or spleen ruptures. The detection of tiny hemorrhages, as often encountered in tumor lesions and often missed in a classical autopsy, may be facilitated. Unenhanced postmortem imaging was so far not able to display these smalltissue lesions with readapted wound margins. Postmortem angiography may therefore detect previously missed lesions as an added value for our discipline.
Postmortem Angiography
A
469
B
C
FIGURE D6.3.17 Selective soft-tissue enhancement that could only insufficiently be reduced by the increase of contrast-agent viscosity demonstrating the centralization of the circulation by maximal dilatation of the precapillary arterioles in these organs at the time of death: (A) Cross-sectional CT image of the brain shows a strong white-matter enhancement, clearly depicting the border between white and gray matter. (B) Short-axis image of the heart shows circular left myocardial enhancement. (C) Coronal lumbar CT image demonstrates distinctive enh ancement of the renal cortex.
As shown within the initial ex vivo porcine model, postmortem angiography based on MRI provides nearly comparable results (Figure D6.3.3). The major drawback of MRI in postmortem angiography is the distinctively longer acquisition time. Especially for whole-body angiography, MSCT is currently faster, although parallel imaging will allow for a fast whole-body MR examination in the near future [152,153] and will soon be evaluated for forensic use. Otherwise, MSCT allows the acquisition of thinner slices down to 0.3 mm that result in an increased quality of the performed 3D reconstructions. It is therefore more likely that MSCT will be more important for postmortem angiography than MRI in the near future because of the lower cost of contrast agents for MSCT and the better accessibility for the forensic discipline. But MRI may serve as an alternative tool for detailed vascular questions when, for instance, calcified plaques and the vascular lumen cannot be distinguished by MSCT. It may solve the problem by selectively enhancing the lumen and thereby distinguishing it from a calcified hypointense plaque. Studies have already shown that MRI is able to further discriminate the different microstructural components of an atherosclerotic plaque, such as lipids or fibrocellular tissue [154–158]. Thereby, lesions that were detected in MSCT can be further investigated using the more elaborate but increased soft-tissue resolutionproviding MRI technique. Postmortem minimally invasive cardiac diagnostics in particular depends on vessel lumen assessment [14,159,160]. Compared with the clinical MSCT assessment of the coronary artery disease, postmortem scans are not influenced by cardiac motion and ventilation and thereby acquire images of increased quality also in angiography examinations. We consider postmortem angiography primarily as a technique for the detection of macroscopic vascular pathologies—such as occlusions, stenosis, or plaques—and we assume that the definitive diagnosis will often require histological investigation of the detected vascular alteration. Therefore, an image-guided biopsy technique that is already about to be implemented [138] will be needed in addition to postmortem angiography to maintain the minimally invasive autopsy approach.
© 2009 by Taylor & Francis Group, LLC
Limitations of the presented approach are currently manifold. The rather low number of subjects in our initial experiments of course requires further studies with a larger spectrum of pathology to prove the reliability of the method. On the other hand, several types of cases will raise problems with the application of the described technique. For example, forensic cases with major vessel injuries, such as seen in fatal hemorrhage, might prevent the required intravascular injection pressure. Cases presenting in an advanced stage of decay might limit the application of the postmortem angiography twofold: Putrefaction gas can lead to intravascular artifacts, and the injection pressure has to be decreased in view of the vulnerability of vessels, predominantly in the mesenteric territory (Figure D6.3.10). However, the first step to a minimally invasive postmortem assessment of the vascular pathology has been made. The anatomy of the arterial system was well displayed, including the intracranial branches as well as the coronary branches. Stenosis and calcified plaques were detected. A soft-tissue injury showed severe contrast-agent extravasation. Therefore the main function of angiography could be successfully achieved.
D6.3.10 ACKNOWLEDGMENTS Many people gave their input to the work that has been condensed in this chapter, but we are particularly grateful to Urs Königsdorfer and Roland Dorn (both Institute of Forensic Medicine, University of Bern) for their experienced support in innumerable ways. Additional thanks go to Therese Perinat (Institute of Forensic Medicine) for the preparation of the tissue specimen and to Susanne Boemke and Kati Haenssgen (both Institute of Human Anatomy, University of Bern) for their reliable collaboration. We would like to express our gratitude to Verena Beutler and Karin Zwygart (Department of Clinical Research, Magnetic Resonance Spectroscopy and Methodology, University of Bern) for their assistance during the data acquisition in MRI. We also thank Andreas Hofer (Medical Technology Department, Inselspital, Bern) for technical support.
470
The Virtopsy Approach
D6.3.11 REFERENCES 1. Thali, M. J., K. Yen, W. Schweitzer, et al. 2003. Virtopsy, a new imaging horizon in forensic pathology: virtual autopsy by postmortem multislice computed tomography (MSCT) and magnetic resonance imaging (MRI)—A feasibility study. J Forensic Sci 48:386–403. 2. Thali, M. J., K. Yen, T. Plattner, et al. 2002. Charred body: virtual autopsy with multi-slice computed tomography and magnetic resonance imaging. J Forensic Sci 47:1326–31. 3. Thali, M. J., W. Schweitzer, K. Yen, et al. 2003. New horizons in forensic radiology: the 60-second digital autopsy-fullbody examination of a gunshot victim by multislice computed tomography. Am J Forensic Med Pathol 24:22–27. 4. Shiotani, S., M. Kohno, N. Ohashi, et al. 2003. Dilatation of the heart on postmortem computed tomography (PMCT): comparison with live CT. Radiat Med 21:29–35. 5. Shiotani, S., M. Kohno, N. Ohashiet al. 2004. Non-traumatic postmortem computed tomographic (PMCT) findings of the lung. Forensic Sci Int 139:39–48. 6. Bisset, R. A., N. B. Thomas, I. W. Turnbull, and S. Lee. 2002. Postmortem examinations using magnetic resonance imaging: four year review of a working service. BMJ 324(7351):1423–24. 7. Patriquin, L., A. Kassarjian, M. Barish, et al. 2001. Postmortem whole-body magnetic resonance imaging as an adjunct to autopsy: preliminary clinical experience. J Magn Reson Imaging 13:277–87. 8. Wallace, S. K., W. A. Cohen, E. J. Stern, and D. T. Reay. 1994. Judicial hanging: postmortem radiographic, CT, and MR imaging features with autopsy confirmation. Radiology 193:263–67. 9. Farkash, U., A. Scope, M. Lynn, et al. 2000. Preliminary experience with postmortem computed tomography in military penetrating trauma. J Trauma 48:303–08. 10. Donchin, Y., A. I. Rivkind, J. Bar-Ziv, J. Hiss, J. Almog, and M. Drescher. 1994. Utility of postmortem computed tomography in trauma victims. J Trauma 37:552–55. 11. Aghayev, E., K. Yen, M. Sonnenschein, et al. 2004. Virtopsy post-mortem multi-slice computed tomograhy (MSCT) and magnetic resonance imaging (MRI) demonstrating descending tonsillar herniation: comparison to clinical studies. Neuroradiology 46:559–64. 12. Jackowski, C., M. Thali, M. Sonnenschein, et al. 2004. Visualization and quantification of air embolism structure by processing postmortem MSCT data. J Forensic Sci 49:1339–42. 13. Thali, M. and P. Vock. 2003. Role of and techniques in forensic imaging. In Forensic Medicine: Clinical and Pathological Aspects, ed. J. Payen-James, A. Busuttil, and W. Smock, 731–45. London: Greenwich Medical Media. 14. Jackowski, C., W. Schweitzer, M. Thali, et al. 2005. Virtopsy: postmortem imaging of the human heart in situ using MSCT and MRI. Forensic Sci Int 149:11–23. 15. Nerantzis, C. E. and P. N. Koutsaftis. 1998. Variant of the left coronary artery with an unusual origin and course: anatomic and postmortem angiographic findings. Clin Anat 11:367–71. 16. Schultz, T. C. 1987. Simple method for demonstrating coronary arteries at postmortem examination. Am J Forensic Med Pathol 8:313–16. 17. Moberg, A. 1967. Anastomoses between extracardiac vessels and coronary arteries. II. Via internal mammary arter-
© 2009 by Taylor & Francis Group, LLC
18.
19.
20.
21.
22.
23.
24.
25.
26.
27.
28.
29.
30.
31.
32.
33.
34.
35.
36.
ies. Post-mortem angiographic study. Acta Radiol Diagn (Stockh) 6:263–72. Nerantzis, C. and D. Avgoustakis. 1980. An S-shaped atrial artery supplying the sinus node area. An anatomical study. Chest 78:274–78. Thomas, A. C. and M. J. Davies. 1985. Post-mortem investigation and quantification of coronary artery disease. Histopathology 9:959–76. Thomas, A. C. and S. Pazios. 1992. The postmortem detection of coronary artery lesions using coronary arteriography. Pathology 24:5–11. Robbins, S. L., M. Solomon, and A. Bennett. 1966. Demonstration of intercoronary anastomoses in human hearts with a low viscosity perfusion mass. Circulation 33:733–43. Weiler, G. and H. J. Knieriem. 1975. Contribution to the morphometry of coronary arteriosclerosis (author’s transl). Z Rechtsmed 75:241–51. Levin, D. C. and J. T. Fallon. 1982. Significance of the angiographic morphology of localized coronary stenoses: histopathologic correlations. Circulation 66:316–20. Rodriguez, F. L. and S. L. Robbins. 1965. Postmortem angiographic studies on the coronary arterial circulation; intercoronary arterial anastomoses in adult human hearts. Am Heart J 70:348–64. Karhunen, P.J., A. Mannikko, A. Penttila, and K. Liesto. 1989. Diagnostic angiography in postoperative autopsies. Am J Forensic Med Pathol 10:303–09. Kauppila, L. I. 1994. Blood supply of the lower thoracic and lumbosacral regions. Postmortem aortography in 38 young adults. Acta Radiol 35:541–44. Segerberg-Konttinen, M. 1987. Demonstration of esophageal varices postmortem by gastroesophageal phlebography. J Forensic Sci 32:703–10. Kauppila, L. I. 1997. Prevalence of stenotic changes in arteries supplying the lumbar spine. A postmortem angiographic study on 140 subjects. Ann Rheum Dis 56:591–95. Karhunen, P. J., A. Penttila, and T. Erkinjuntti. 1990. Arteriovenous malformation of the brain: imaging by postmortem angiography. Forensic Sci Int 48:9–19. Karhunen, P. J. 1991. Neurosurgical vascular complications associated with aneurysm clips evaluated by postmortem angiography. Forensic Sci Int 51:13–22. Weman, S. M., U. S. Salminen, A. Penttila, A. Mannikko, and P. J. Karhunen. 1999. Post-mortem cast angiography in the diagnostics of graft complications in patients with fatal outcome following coronary artery bypass grafting (CABG). Int J Legal Med 112:107–14. Karhunen, P. J., R. Kauppila, A. Penttila, and T. Erkinjuntti. 1990. Vertebral artery rupture in traumatic subarachnoid haemorrhage detected by postmortem angiography. Forensic Sci Int 44:107–15. Karhunen, P. J. and A. Penttila. 1989. Diagnostic postmortem angiography of fatal splenic artery haemorrhage. Z Rechtsmed 103:129–36. Karhunen, P. J. and A. Servo. 1993. Sudden fatal or nonoperable bleeding from ruptured intracranial aneurysm. Evaluation by post-mortem angiography with vulcanising contrast medium. Int J Legal Med 106:55–59. Kauppila, L. I. 1995. Ingrowth of blood vessels in disc degeneration. Angiographic and histological studies of cadaveric spines. J Bone Joint Surg Am 77:26–31. Saimanen, E., A. Jarvinen, and A. Pentitila. 2001. Cerebral cast angiography as an aid to medicolegal autopsies in
Postmortem Angiography
37.
38.
39.
40.
41.
42.
43. 44.
45.
46.
47.
48.
49.
50. 51.
52.
53.
54.
55.
cases of death after adult cardiac surgery. Int J Legal Med 114:163–68. Smith, M., D. E. Trummel, M. Dolz, and S. J. Cina. 1999. A simplified method for postmortem coronary angiography using gastrograffin. Arch Pathol Lab Med 123:885–88. Prahlow, J. A., E. S. Scharling, and P. E. Lantz. 1996. Postmortem coronary subtraction angiography. Am J Forensic Med Pathol 17:225–30. Russell, G. A. and P. J. Berry. 1988. Post mortem radiology in children with congenital heart disease. J Clin Pathol 41: 830–36. Rah, B. R., R. J. Katz, A. G. Wasserman, and J. S. Reiner. 2001. Post-mortem three-dimensional reconstruction of the entire coronary arterial circulation using electron-beam computed tomography. Circulation 104:3168. Dirnhofer, R., C. Jackowski, P. Vock, K. Potter, and M. J. Thali. 2006. VIRTOPSY: minimally invasive, imagingguided virtual autopsy. Radiographics 26:1305–33. Jackowski, C., M. Sonnenschein, M. J. Thali, et al. 2005. Virtopsy: postmortem minimally invasive angiography using cross section techniques—implementation and preliminary results. J Forensic Sci 50:1175–86. Parade, G. W. 1933. Coronardarstellung. Verh Dtsch Ges Inn Med 45:216–20. Miyata, S. 1939. Aufbau und Gestalt der peripheren arteriellen Strombahn des kleinen Kreislaufs. Virchows Arch Path Anat 304:608–24. Hintze, A. 1933. Fehlbildungen im Blutgefässsystem und ihr Nachweis mittels der Röntgenuntersuchung. Virchows Arch Pathol Anat 289:705–17. Schlesinger, M. J. 1938. An injection plus dissection study of coronary artery occlusion and anastomosis. Am Heart J 15:528–68. Hinman, F. and D. M. Morison. 1924. Comparative study of circulatory changes in hydronephrosis, caseo-cavernosous tuberculosis, and polycstic kidney. A preliminary report. J Urol 11:131–41. Crainicianu, A. 1922. Anatomische Studien über die Coronararterien und experimentelle Untersuchungen über ihre Durchgängigkeit. Virchows Arch 238:1–75. van Dantzig, J. M. and A. E. Becker. 1986. Sudden cardiac death and acute pathology of coronary arteries. Eur Heart J 7:987–91. Dock, W. 1941. The capacity of the coronary bed in cardiac hypertrophy. J Exp Med 74:177–86. Falk, E. 1985. Unstable angina with fatal outcome: dynamic coronary thrombosis leading to infarction and/or sudden death. Autopsy evidence of recurrent mural thrombosis with peripheral embolization culminating in total vascular occlusion. Circulation 71:699–708. Farb, A., A. L. Tang, A. P. Burke, L. Sessums, Y. Liang, and R. Virmani. 1995. Sudden coronary death. Frequency of active coronary lesions, inactive coronary lesions, and myocardial infarction. Circulation 92:1701–09. Freudenberg, V. H., H. J. Knieriem, C. Moller, and C. Janzen. 1974. [Quantitative morphology of coronary arteriosclerosis and coronary insufficiency, author’s translation]. Basic Res Cardiol 69:161–203. Galbraith, J. E., M. L. Murphy, and N. de Soyza. 1978. Coronary angiogram interpretation. Interobserver variability. J Am Med Assoc 240:2053–56. Hochman, J. S., W. J. Phillips, D. Ruggieri, and S. F. Ryan. 1988. The distribution of atherosclerotic lesions in the coro-
© 2009 by Taylor & Francis Group, LLC
471
56.
57.
58. 59.
60.
61.
62. 63.
64. 65.
66.
67.
68.
69.
70.
71. 72.
73.
nary arterial tree: relation to cardiac risk factors. Am Heart J 116(5 Pt 1):1217–22. Hutchins, G. M., B. H. Bulkley, R. L. Ridolfi, L. S. Griffith, F. T. Lohr, and M. A. Piasio. 1977. Correlation of coronary arteriograms and left ventriculograms with postmortem studies. Circulation 56:32–37. Lagundoye, S. B., G. M. Edington, G. Ibeachum, and W. P. Cockshott. 1976. Post mortem coronary arteriography in Nigerians: a radiological review. Afr J Med Med Sci 5:9–17. Laurie, W. and J. D. Woods. 1962. Interarterial coronary anastomoses in three race groups. Lancet 1:13–17. Pepler, W. J. and B. J. Meyer. 1960. Interarterial coronary anastomoses and coronary arterial pattern. A comparative study of South African Bantu and European hearts. Circulation 22:14–24. Plachta, A., S. A. Thompson, and F. D. Speer. 1955. Pericardial and myocardial vascularization following cardiopericardiopexy; magnesium silicate technique. AMA Arch Pathol 59:151–61. Prinzmetal, M., S. Kayland, C. Margoles, and L. J. Tragerman. 1942. A quantitative method for determing collateral coronary circulation. J Mt Sinai Hosp N Y 8:933–45. Rissanen, V. T. 1970. Double contrast technique for postmortem coronary angiography. Lab Invest 23:517–20. Rozenberg, V. D. and L. M. Nepomnyashchikh. 2002. Pathomorphology of myocardial bridges and their role in the pathogenesis of coronary disease. Bull Exp Biol Med 134:593–96. Schoenmackers, J. 1965. Die Angiomorphologie der Koronarangiogramme. Roentgenfortschritte 102:349–68. Schoenmackers, J., F. J. Bultmann, and G. Dechene. 1974. Comparative planimetric and stereoscopic analysis of postmortal coronarogrammes. Z Kardiol 63:689–97. Trask, N., R. M. Califf, M. J. Conley, et al. 1984. Accuracy and interobserver variability of coronary cineangiography: a comparison with postmortem evaluation. J Am Coll Cardiol 3:1145–54. Weiler, G., K. D. Erkrath, and M. Risse. 1979. [Contribution to anastomotic coronary circulation illustrated by a BlandWhite-Garland-syndome in an adult, author’s translation]. Z Kardiol 68:717–19. Allison, R. B., F. L. Rodriguez, E. A. Higgins, Jr., et al. 1963. Clinicopathologic correlations in coronary atherosclerosis. Four hundred thirty patients studied with postmortem coronary angiography. Circulation 27:170–84. Barmeyer, J. 1968. Post mortem coronary angiography and perfusion of normal and diseased hearts, perfusibility of intercoronary anastomoses. Beitr Pathol Anat 137: 373–90. Barmeyer, J. and H. Reindell. 1971. Intercoronary anastomoses in postmortum angiography. Radiologe 11: 198–202. Bellman, S. and H. A. Frank. 1958. Intercoronary collaterals in normal hearts. J Thorac Surg 36:584–603. Bulkely, B. H. and G. M. Hutchins. 1977. Myocardial consequences of coronary artery bypass graft surgery. The paradox of necrosis in areas of revascularization. Circulation 56:906–13. Coghill, S.B., S. M. Nicoll, A. McKimmie, I. Houston, and B. M. Matthew. 1983. Revitalising postmortem coronary angiography. J Clin Pathol 36:1406–09.
472
74. Davis, N. A. 1963. A radioisotope dilution technique for the quantitative study of coronary artery disease postmortem. Lab Invest 12:1198–203. 75. Estes, E.H., Jr., M. L. Entman, H.B. Dixon, and D. B. Hackel. 1966. The vascular supply of the left ventricular wall. Anatomic observations, plus a hypothesis regarding acute events in coronary artery disease. Am Heart J 71:58–67. 76. Giraldo, A. A., M. J. Higgins, and J. J. Humes. 1986. Anatomical methods in the study of cardiovascular pathology: a refined technique. Ann Clin Lab Sci 16:13–25. 77. Giraldo, A. A. and M. J. Higgins. 1988. Laboratory methods in the study of coronary atherosclerosis. Pathol Annu 23 (Pt 1):217–36. 78. Gray, C. R., H. A. Hoffman, W. S. Hammond, K. L. Miller, and R. O. Oseasohn. 1962. Correlation of arteriographic and pathologic findings in the coronary arteries in man. Circulation 26:494–99. 79. McNamara, J. J., M. A. Molot, J. F. Stremple, and R. T. Cutting. 1971. Coronary artery disease in combat casualties in Vietnam. J Am Med Assoc 216:1185–87. 80. Müller-Mohnssen, H. 1957. Topography of septal arteries in the human heart and its significance in the origin of collateral circulation in coronary sclerosis. Beitr Pathol Anat 118:121–42. 81. Murphy, M. L., J. E. Galbraith, and N. de Soyza. 1979. The reliability of coronary angiogram interpretation: an angiographic-pathologic correlation with a comparison of radiographic views. Am Heart J 97:578–84. 82. Nerantzis, C. E. and P. N. Koutsaftis. 1998. Variant of the left coronary artery with an unusual origin and course: anatomic and postmortem angiographic findings. Clin Anat 11:367–71. 83. Nerantzis, C. E., C. A. Lefkidis, T. B. Smirnoff, E. B. Agapitos, and P. S. Davaris. 1998. Variations in the origin and course of the posterior interventricular artery in relation to the crux cordis and the posterior interventricular vein: an anatomical study. Anat Rec 252:413–17. 84. Nerantzis, C. E. and S. K. Marianou. 2000. Ectopic “high” origin of both coronary arteries from the left aortic wall: anatomic and postmortem angiographic findings. Clin Anat 13:383–86. 85. Oosawa, K. 1967. Clinicopathological studies on the coronary circulation with postmortem coronary angiography. Gunma J Med Sci 16:146–76. 86. Schwartz, J. N., Y. Kong, D. B. Hackel, and A. G. Bartel. 1975. Comparison of angiographic and postmortem findings in patients with coronary artery disease. Am J Cardiol 36:174–78. 87. Eusterman, J. H., R. W. P. Achor, O. W. Kincaid, and A. L. Brown. 1962. Atherosclerotic disease of the coronary arteries—a pathologic-radiologic correlative study. Circulation 26:1288–95. 88. Farrer-Brown, G. and P. M. Rowles. 1969. Vascular supply of interventricular septum of human heart. Br Heart J 31:727–34. 89. Fulton, W. F. 1963. Arterial anastomoses in the coronary circulation. I. Anatomical features in normal and diseased hearts demonstrated by stereoarteriography. Scott Med J 143:420–34. 90. Giese, W. 1957. Anastomoses of coronary arteries in coronary arteriosclerosis. Dtsch Med Wochenschr 82:602–04. 91. Heard, B. E. 1976. Pathology of hearts after aortocoronary saphenous vein bypass grafting for coronary artery disease, studied by post-mortem coronary angiography. Br Heart J 38:838–59.
© 2009 by Taylor & Francis Group, LLC
The Virtopsy Approach
92. Jarvinen, A., A. Manniko, P. Ketonen, M. SegerbergKonttinen, and R. Luosto. 1989. Surgical technique and operative mortality in coronary artery bypass. A postmortem analysis with castangiography. Scand J Thorac Cardiovasc Surg 23:103–09. 93. Vesterby, A. 1981. Postmortem coronary angiography and histological investigation of the conduction system of the heart in sudden unexpected death due to coronary heart disease. Acta Pathol Microbiol Scand [A] 89:157–63. 94. Weman, S. M., P. J. Karhunen, A. Penttila, A. A. Jarvinen, and U. S. Salminen. 2000. Reperfusion injury associated with one-fourth of deaths after coronary artery bypass grafting. Ann Thorac Surg 70:807–12. 95. Melnick, G. S., N. Tuna, and M. J. Gilson. 1963. Postmortem coronary arteriogram. A correlation with electrocardiographic and anatomic findings. Angiology 14: 252–59. 96. van der Straten, P. P. 1955. La Coronarographie post mortem de l´homme age. Acta Cardiol 10:15–43. 97. Brascho, D. J. 1963. A technique for postmortem coronary arteriography. Am Heart J 66:375–80. 98. Huguet, J. F., R. Luccioni, B. Navarro, and J. Colonna. 1970. Coronary anastomoses. Post-mortem radiographic study. Ann Radiol (Paris) 13:651–66. 99. Kohlhardt, M., H. Müller-Marienburg, G. Vita, and E. Zeitler. 1963. Determination of the coincidence between coronarography and pathologico-anatomical findings. Fortschr Geb Rontgenstr Nuklearmed 98:399–408. 100. Ravin, A. and Geever, E. F. 1946. Coronary arteriosclerosis, coronary anastomoses and myocardial infarction. Arch Intern Med 78:125–38. 101. Wexberg, P., M. Gottsauner-Wolf, I. Sulzbacher, P. Birner, A. Laggner, and D. Glogar. 2001. Fatal late coronary thrombosis after implantation of a radioactive stent: postmortem angiographic and histologic findings—case report. Radiology 220:142–44. 102. Rodriguez, F. L. and S. L. Robbins. 1965. Postmortem angiographic studies on the coronary arterial circulation; intercoronary arterial anastomoses in adult human hearts. Am Heart J 70:348–64. 103. Weitzman, D. 1964. Post-mortem coronary arteriography and its correlation with electrocardiography. Br Heart J 26:330–36. 104. Fujikura, T., Y. Kominato, I. Shimada, N. Hata, and H. Takizawa. 1990. Forensic application of angiography on injured brain. Med Sci Law 30:127–32. 105. Kormano, M. and K. Reijonen. 1973. Microangiographic filling of the vascular system of the brain. Neuroradiology 6:83–86. 106. Saunders, R. L. de C. H. 1960. Microangiography of the brain and spinal cord. In X-Ray Microscopy and X-Ray Microanalysis, Proceedings of the Second International Symposium held in Stockholm, Sweden, 1960, ed. A. Engström, V. Cosslett, and H. Pattee, 244–56. Amsterdam: Elsevier. 107. Maxeiner, H. 2001. Demonstration and interpretation of bridging vein ruptures in cases of infantile subdural bleedings. J Forensic Sci 46:85–93. 108. Maxeiner, H. 1997. Detection of ruptured cerebral bridging veins at autopsy. Forensic Sci Int 89:103–10. 109. Hassler, O. 1966. Deep cerebral venous system in man. A microangiographic study on its areas of drainage and its anastomoses with the superficial cerebral veins. Neurology 16:505–11.
Postmortem Angiography
110. Ehrlich, E., H. Maxheiner, and J. Lange. 2004. Postmortem radiological investigation of bridging vein ruptures. Legal Med 5:S225–27. 111. Dowling, G. and B. Curry. 1988. Traumatique basal subarachnoid hemorrhage. Report of six cases and review of the literature. Am J Forensic Med Pathol 9:23–31. 112. Dor, P. and G. Salamon. 1970. The arterioles and capillaries of the brain stem and cerebellum: a microangiographic study. Neuroradiology 1:27–29. 113. Mant, A. K. 1972. Traumatic subarachnoid haemorrhage following blows to the neck. J Forensic Sci Soc 12: 567–72. 114. Stein, B. M., W. McCormick, J. N. Rodriguez, and J. M. Taveras. 1961. Incidence and significance of occlusive vascular disease of the extracranial arteries as demonstrated by post-mortem angiography. Trans Am Neurol Assoc 86:60–66. 115. Stein B. M. and G. T. Svare. 1963. A technique of postmortem angiography for evaluating arteriosclerosis of the aortic arch and carotid and vertebral arteries. Radiology 81:252–56. 116. Charr, R. and J. Wooddrow Savacool. 1940. Changes in the arteries in the walls of tuberculous pulmonary cavities. Arch Pathol 30:1159–71. 117. Milne, E. N. 1967. Circulation of primary and metastatic pulmonary neoplasms. A postmortem microarteriographic study. Am J Roentgenol Radium Ther Nucl Med 100:603–19. 118. Gloor, F. 1953. Vascularization of the esophagus. Thoraxchirurgie 1:146–67. 119. Faller, A. 1957. Behavior of arterial vessels in various parietal layers of the rectum in man. Acta Anat (Basel) 30:275–86. 120. Reiner, L., F. L. Rodriguez, R. Platt, and M. J. Schlesinger. 1959. Injection studies on the mesenteric arterial circulation. I. Technique and observations on collaterals. Surgery 45:820–33. 121. Aranyi, Z., M. Patanyik, G. Nemeth, M. Scholz, and J. Stumpf. 1985. Post mortem uterine arteriography and in vivo angiographic diagnosis. Acta Morphol Hung 33:219–25. 122. Kauppila, L. I., P. J. Karhunen, and U. Lahdenranta. 1994. Intermittent medullary claudication: postmortem spinal angiographic findings in two cases and in six controls. J Spinal Disord 7:242–47. 123. Dejdar, R., H. Roubkova, M. Cachovan, J. Kruml, and J. Linhart. 1967. Comparison of postmortem angiograms with macro and microscopic findings on the A. femoralis and A. poplitea. Arch Kreislaufforsch 54:309–35. 124. Ross C. F. and K. D. Keele. 1951. Post mortem arteriography in “normal” lower limbs. Angiology 2:374–85. 125. Gronvall, J. and N. Graem. 1989. Radiography in postmortem examinations of fetuses and neonates. Findings on plain films and at arteriography. APMIS 97:274–80. 126. Richter, E. 1976. Postmortem angiocardiography in newborn infants with congenital malformation of the heart and great vessels. Pediatr Radiol 4:133–38. 127. Foote, G. A., A. J. Wilson, and J. H. Stewart. 1978. Perinatal post-mortem radiography—experience with 2500 cases. Br J Radiol 51:351–56. 128. Stoeter, P., D. Petersen, and K. Voigt. 1977. [Comparative roentgenological embryology of the supraaortic arteries of domestic mammals with a prenatal angiographic technique, authors’ translation]. Rofo 126:150–56. 129. Jeanmart, L. 1974. Study of the cerebral vascularization of the human fetus by transumbilical angiography. Angiology 25:334–49.
© 2009 by Taylor & Francis Group, LLC
473
130. Beck, B. L. 1987. Two cases of congenital vascular malformations proved by post-mortem arteriography. Case reports. Acta Pathol Microbiol Immunol Scand [A] 95:17–21. 131. Potts, D.G., G. T. Svare, and R. T. Bergeron. 1969. The developing brain. Correlation between radiologic and anatomical findings. Acta Radiol Diagn (Stockh) 9:430–39. 132. Grabuschnigg, P. and F. Rous. 1990. Preservation of human cadavers throughout history—a contribution to development and methodology. Beitr Gerichtl Med 48:455–58. 133. Hanzlick, R. 1994. Embalming, body preparation, burial, and disinterment. An overview for forensic pathologists. Am J Forensic Med Pathol 15:122–31. 134. Macdonald, G. J. and D. B. MacGregor. 1997. Procedures for embalming cadavers for the dissecting laboratory. Proc Soc Exp Biol Med 215:363–65. 135. Brown, K. T., G. I. Getrajdman, and J. F. Botet. 1995. Clinical trial of the Bard CT guide system. J Vasc Interv Radiol 6:405–10. 136. Fichtinger, G., T. L. DeWeese, A. Patriciu, et al. 2002. System for robotically assisted prostate biopsy and therapy with intraoperative CT guidance. Acad Radiol 9:60–74. 137. Masamune, K., G. Fichtinger, A. Patriciu, et al. 2001. System for robotically assisted percutaneous procedures with computed tomography guidance. Comput Aided Surg 6:370–83. 138. Aghayev, E., M. J. Thali, M. Sonnenschein, C. Jackowski, R. Dirnhofer, and P. Vock. 2007. Post-mortem tissue sampling using computed tomography guidance. Forensic Sci Int 166:199–203. 139. Bohm, E. 1983. Results of postmortem organ and tissue perfusions. Beitr Gerichtl Med 41:449–58. 140. Stoeter, P., M. Buchhocker, W. Bruzek, U. Drews, and K. Schulze. 1980. [Angiographic examinations of the circulatory development of living chick embryos, authors’ translation]. Rofo 133:83–91. 141. Bohm, E. 1983. A preparative technique for morphological analysis of the vessels in head and neck for medico-legal examinations. Scan Electron Microsc (Pt 4):1973–81. 142. Bohm, E. and F. Hubner. 1983. Microradiographic findings in death by hanging. Beitr Gerichtl Med 41:465–73. 143. Hübner, F. and E. Böhm. 1985. Zur forensischen Bedeutung postmortaler Injektions- und Perfusionstechniken. Zacchia 48:95–120. 144. Rozenberg, V. D. 1987. Postmortem contrast cardioventriculography. Arkh Patol 49:82–84. 145. Rubin, P., G. W. Casarett, S. S. Kurohara, and M. Fujii. 1964. Microangiography as a technique; radiation effect versus artifact. Am J Roentgenol Radium Ther Nucl Med 92:378–87. 146. Bergquist, E. and L. Rammer. 1974. Postmortem vertebral angiography in cases of traumatic subarachnoid hemorrhage. Radiology 110:709–10. 147. Rozenberg, V. D. 1989. Pathomorphological data of differential diagnosis of dilatational cardiomyopathy. Vrach Delo 12:18–20. 148. Frik, W. and W. F. Persch. 1969. The effect of contrast media type on the vascular caliber in experimental angiography. Fortschr Geb Rontgenstr Nuklearmed 111:620–29. 149. Schoenmackers, J. 1960. Technik der postmortalen Angiographie mit Berücksichtigung verwandter Methoden postmortaler Gefässdarstellung. Ergebn allg Pathol Anat 39:53–151.
474
150. Jackowski, C., S. Bolliger, E. Aghayev, et al. 2006. Reduction of postmortem angiography-induced tissue edema by using polyethylene glycol as a contrast agent dissolver. J Forensic Sci 51:1134–37. 151. Kasser, U., H. Kroemer, G. Altrock, and P. Heimburg. 1988. Reference ranges of viscoelasticity of human blood. Biorheology 25:727–41. 152. Quick, H. H., F. M. Vogt, S. Maderwald, et al. 2004. High spatial resolution whole-body MR angiography featuring parallel imaging: initial experience. Rofo 176:163–69. 153. Kramer, H., S. O. Schoenberg, K. Nikolaou, et al. 2004. Cardiovascular whole body MRI with parallel imaging. Radiologe 44:835–43. 154. Toussaint, J. F., G. M. LaMuraglia, J. F. Southern, V. Fuster, and H. L. Kantor. 1996. Magnetic resonance images lipid, fibrous, calcified, hemorrhagic, and thrombotic components of human atherosclerosis in vivo. Circulation 94:932–38. 155. Wasserman, B. A., W. I. Smith, H. H. Trout, III, R. O. Cannon, III, R. S. Balaban, and A. E. Arai. 2002. Carotid artery atherosclerosis: in vivo morphologic characterization with gadolinium-enhanced double-oblique MR imaging initial results. Radiology 223:566–73.
© 2009 by Taylor & Francis Group, LLC
The Virtopsy Approach
156. Ruehm, S. G. 2003. Magnetic resonance imaging of atherosclerotic plaque. Herz 28:513–20. 157. Worthley, S. G., G. Helft, V. Fuster, et al. 2000. Noninvasive in vivo magnetic resonance imaging of experimental coronary artery lesions in a porcine model. Circulation 101:2956–61. 158. Worthley, S. G., G. Helft, V. Fuster, et al. 2000. High resolution ex vivo magnetic resonance imaging of in situ coronary and aortic atherosclerotic plaque in a porcine model. Atherosclerosis 150:321–29. 159. Jachau, K., T. Heinrichs, W. Kuchheuser, et al. 2004. Computed tomography and magnetic resonance imaging compared to pathoanatomic findings in isolated human autopsy hearts. Rechtsmedizin 14:109–16. 160. Jackowski, C., A. Christe, M. Sonnenschein, E. Aghayev, and M. J. Thali. 2006. Postmortem unenhanced magnetic resonance imaging of myocardial infarction in correlation to histological infarction age characterization. Eur Heart J 27:2459–67. 161. Jackowski, C., Persson, A., and Thali, M. J. Whole body postmortem angiography with a high viscosity contrast agent solution using polyethylene glycol as contrast agent dissolver. J Forensic Sci 2008 Mar; 53(2): 465–8.
D7
Experiences with Virtual Autopsy Approach Worldwide Lars Oesterhelweg and Michael J. Thali
CONTENTS D7.1 D7.2 D7.3 D7.4
Asia ................................................................................................................................................................................ 475 Australia......................................................................................................................................................................... 476 Europe ............................................................................................................................................................................ 476 North America ............................................................................................................................................................... 477
As far as we know, only a few forensic institutes worldwide have research projects in forensic radiology or are using modern imaging techniques routinely. In this chapter, major publications and working groups are presented by continents and in alphabetical order of their countries.
D7.1 ASIA Early attempts in the postmortem use of computed tomography were performed in Israel on trauma victims. First publications are from the early 1990s [1–5]. In Japan, an Autopsy Imaging Association was founded a few years ago. Many scientific articles were published in the field of postmortem use of CT, magnetic resonance imaging (MRI), and ultrasound and in the processing of achieved data [6–24]. In addition, the first studies from the People’s Republic of China have recently been published [25,26]. 1. Donchin, Y., A. I. Rivkind, J. Bar-Ziv, J. Hiss, J. Almog, and M. Drescher. 1994. Utility of postmortem computed tomography in trauma victims. J Trauma 37:552–26. 2. Farkash, U., A. Scope, M. Lynn, et al. 2000. Preliminary experience with postmortem computed tomography in military penetrating trauma. J Trauma 48:303–09. 3. Kahana, T. and J. Hiss. 1999. Forensic radiology. Br J Radiol 72:129–33. 4. Kahana, T. and J. Hiss. 1997. Identification of human remains: forensic radiology. J Clin Forensic Med 4:7–15. 5. Kahana, T., J. A. Ravioli, C. L. Urroz, and J. Hiss. 1997. Radiographic identification of fragmentary human remains from a mass disaster. Am J Forensic Med Pathol 18:40–44. 6. Arai, A., S. Shiotani, K. Yamazaki, et al. 2006. Postmortem computed tomographic (PMCT) and postmortem magnetic resonance imaging (PMMRI) demonstration of fatal massive retroperitoneal hemorrhage caused by abdominal aortic aneurysm (AAA) rupture. Radiat Med 24:147–49. 7. Ezawa H., S. Shiotani, and S. Uchigasaki. 2007. Autopsy imaging in Japan. Rechtsmedizin 17:19–20. 8. Ezawa, H., R. Yoneyama, S. Kandatsu, K. Yoshikawa, H. Tsujii, and K. Harigaya. 2002. Autopsy imaging (AI)—a new concept for autopsy. Byori to Rinsho 20:633–41.
9. Ezawa, H., R. Yoneyama, S. Kandatsu, K. Yoshikawa, H. Tsujii, and K. Harigaya. 2003. Introduction of autopsy imaging redefines the concept of autopsy: 37 cases of clinical experience. Pathology Int 53:865–73. 10. Ezawa, H., R. Yoneyama, S. Kandatsu, K. Yoshikawa, H. Tsujii, and K. Harigaya. 2000. The constitution of autopsy imaging system (AIS). J Japan Path Assoc 90:178. 11. Hamano, J., S. Shiotani, K. Yamazaki, M. Suzuki, and H. Ishikawa. 2004. Postmortem computed tomographic (PMCT) demonstration of fatal hemoptysis by pulmonary tuberculosis—radiological–pathological correlation in a case of rupture of Rasmussen’s aneurysm. Radiat Med 22:120–22. 12. Hayakawa, M., S. Yamamoto, H. Motani, D. Yajima, Y. Sato, and H. Iwase. 2006. Does imaging technology overcome problems of conventional postmortem examination? A trial of computed tomography imaging for postmortem examination. Int J Leg Med 120:24–26. 13. Ikeda, G., R. Yamamoto, M. Suzuki, H. Ishikawa, K. Kikuchi, and S. Shiotani. 2007. Postmortem computed tomography and magnetic resonance imaging in a case of terminal-stage small cell lung cancer: an experience of autopsy imaging in tumor-related death. Radiat Med 25:84–87. 14. Shiotani, S., M. Kohno, N. Ohashi, et al. 2003. Dilatation of the heart on postmortem computed tomography (PMCT): Comparison with live CT. Radiation Med 21:29–35. 15. Shiotani, S., M. Kohno, N. Ohashi, et al. 2004. Nontraumatic postmortem computed tomographic (PMCT) findings of the lung. Forensic Sci Int 139:39–48. 16. Shiotani, S., K. Yamazaki, K. Kikuchi, et al. 2005. Postmortem magnetic resonance imaging (PMMRI) demonstration of reversible injury phase myocardium in a case of sudden death from acute coronary plaque change. Radiat Med 23:563–65. 17. Suzuki, N., A. Takatsu, A. Shigeta, O. Kitamura, and S. Murata. 1996. Application of 3D and 4D image analysis of forensic medicine. Current Topics in Forensic Sciences 1:94–96. 18. Suzuki, N., A. Takatsu, A. Shigeta, O. Kitamura, and S. Murata. 1996. Introduction of virtual reality to forensic pathology. Current Topics in Forensic Sciences 1:115-7. 19. Takatsu, A., N. Suzuki, A. Hattori, and A. Shigeta. 1999. The concept of the digital morgue as a 3D database. Legal Med 1:29–33. 475
© 2009 by Taylor & Francis Group, LLC
476
The Virtopsy Approach
20. Takatsu, A., N. Suzuki, A. Hattori, A. Shigeta, and S. Abe. 2007. High-dimensional medical imaging and virtual reality techniques. Development of the advanced digital morgue. Rechtsmedizin 17:13–18. 21. The Japan Society of Autopsy Imaging (AI). 2007. Homepage: http://plaza.umin.ac.jp/%7Eai-ai/english.htm (accessed September 18, 2007). 22. Uchigasaki, S., L. Oesterhelweg, A. Gehl, et al. 2004. Application of compact ultrasound imaging device to postmortem diagnosis. Forensic Sci Int 140:33–41. 23. Uchigasaki, S., L. Oesterhelweg, J. P. Sperhake, K. Püschel, and S. Oshida. 2006. Application of ultrasonography to postmortem examination. Diagnosis of pericardial tamponade. Forensic Sci Int 162:167–69. 24. Yoshino, M., M. Taniguchi, K. Imaizumi, et al. 2005. A new retrieval system for a database of 3D facial images. Forensic Sci Int 148:113–20. 25. Hao, Z. R., J. D. Wu, X. S. Liu, B. Z. Chen, T. Hu, and H. W. Xing. 2007. Application of virtopsy in forensic science. Fa Yi Xue Za Zhi 23:142–44. 26. Xie, Y., X. F. Yi, X. G. Cheng, et al. 2006. The application of radiological imaging in the forensic pathology about cervical part. Fa Yi Xue Za Zhi 22:378–80.
D7.2 AUSTRALIA Since 2004, the Victorian Institute of Forensic Medicine (VIFM) in Melbourne has used an MSCT scanner for postmortem investigations. This scanner is also used by the Center of Human Identification (CHI) at the VIFM. [1] 1. Blau, S., S. Robertson, and M. Johnstone. 2008. Disaster victim identification: New applications for postmortem computed tomography. J Forensic Sci 53:956–961.
D7.3 EUROPE At the University of Toulouse in France, F. Dedouit and colleagues are working toward implementing modern imaging techniques in postmortem investigations and have published several articles [1–6]. In German-speaking nations, besides the Institute of Forensic Medicine in Bern, Switzerland, scientific studies on the use of forensic radiology have been performed by the Institute of Legal Medicine at the University of Heidelberg (K. M. Stein) and by the Institute of Legal Medicine at the University of Cologne (T. Riepert). Also the Institute of Legal Medicine at the University of Schlesing-Holstein has presented studies on the use of postmortem imaging in gunshot victims. In other institutes of legal medicine such as at the University of Hamburg and Ulm, research in the field of forensic imaging is beginning [7–15]. In Scandinavia to date there are two institutes that deal on a routine basis with modern imaging techniques. At the Institute of Forensic Medicine in Copenhagen, Denmark, since 2001 a CT scanner has been employed, and in 2004 a MRI scanner was acquired. K. Poulsen and J. Simonsen gave the major publication in 2007. Very promising study designs also come from the Center for Medical Image Science and Visualization (CMIV) (A. Persson) in close cooperation with © 2009 by Taylor & Francis Group, LLC
the Institute of Forensic Medicine at the Linköping University in Sweden [16–21]. In the United Kingdom, postmortem imaging activities exist and postmortem MRI studies are published. Attempts to establish imaging techniques in forensic pathology are supported by G. N. Rutty, head of the Institute of Forensic Medicine in Leicester [22–39]. 1. Dedouit, F., F. Loubes-Lacroix , R. Costagliola, et al. 2007. Post-mortem changes of the middle ear: Multislice computed tomography study. Forensic Sci Int 175:149–54. 2. Dedouit, F., P. Otal, R. Costagliola, et al. 2006. Application à la thanatologie de l’imagerie en coupe: revue iconographique. J Radiol 87:619–38. 3. Dedouit, F., P. Otal, F. Loubes-Lacroix, et al. 2006. Posttraumatic venous and systemic air embolism associated with spinal epidural emphysema: multi-slice computed tomography diagnosis. Forensic Sci Int 158:190–94. 4. Dedouit, F., N. Telmon, R. Costagliola, P. Otal, F. Joffre, and D. Rougé. 2007. Virtual anthropology and forensic identification: Report of one case. Forensic Sci Int 173:182–87. 5. Dedouit, F., G. Tournel, A. Bécart, V. Hédouin, and D. Gosset. 2007. Suicidal hanging resulting in complete decapitation—forensic, radiological, and anthropological studies: a case report. J Forensic Sci 52(5):1190–93. 6. Dedouit, F., N. Telmon, C. Guilbeau-Frugier, et al. 2007. Virtual autopsy and forensic identification—practical application: a report of one case. J Forensic Sci 52:960-4. 7. Oesterhelweg, L., M. Lorenzen, C. Braun, D. Rohwedder, G. Adam, and K. Püschel. 2007 Radiosektion—CT-assistierte Rekonstruktion eines erweiterten Suizids. Rechtsmedizin 17:44–47. 8. Paperno, S., T. Riepert, B. Krug, et al. 2005. Value of postmortem computed tomography in comparison to autopsy. Rofo 177:130–36. 9. Pohlenz, P., M. Blessmann, L. Oesterhelweg, et al. 2007. 3D C-arm as an alternative modality to CT in postmortem imaging: Technical feasibility. Forensic Sci Int 175:134–39. 10. Püschel, K. 2007. Schöne neue Welt von Virtopsy®, Autopsy Imaging, Radiosektion und Nekroradiologie. Rechtsmedizin 17:5–6. 11. Riepert, T., C. Rittner, D. Ulmcke, S. Ogbuihi, and F. Schweden. 1995. Identification of an unknown corpse by means of computed tomography (CT) of the lumbar spine. J Forensic Sci 40:126–27. 12. Riepert, T., D. Ulmcke, U. Jendrysiak, and C. Rittner. 1995. Computer-assisted simulation of conventional roentgenograms from three-dimensional computed tomography (CT) data—an aid in the identification of unknown corpses (FoXSIS). Forensic Sci Int 71:199–204. 13. Stein, K. M., M. L. Bahner, J. Merkel, S. Ain, and R. Mattern. 2000. Detection of gunshot residues in routine CTs. Int J Legal Med 114:15–18. 14. Stein, K. M., K. Ruf, M. K. Ganten, and R. Mattern. 2006. Representation of cerebral bridging veins in infants by postmortem computed tomography. Forensic Sci Int 163:93–101. 15. Oehmichen. M., H.B. Gehl, C. Meissner, et al. 2003. Forensic pathological aspects of postmortem imaging of gunshot injury to the head: Documentation and biometric data. Acta Neuropathol 105:570–580. 16. Center for Medical Image Science and Visualization (CMIV). 2007. Homepage (list of publications). http:// www.cmiv.liu.se/output/publications (accessed September 18, 2007).
Experiences with Virtual Autopsy Approach Worldwide
17. Ljung, P., C. Winskog, A. Persson, C. Lundström, and A. Ynnerman. 2006. Full body virtual autopsies using a stateof-the-art volume rendering pipeline. Trans Visualization Computed Graphics 12:869–76. 18. Persson, A., C. Jackowski, E. Engstrom, and H. Zachrisson. 2008. Advances of dual source, dual energy imaging in post-mortem CT. Eur J Radiol, in press, doi: 10.1016/j. ejrad.2008.05.008. 19. Jackowski, C., A. Persson, M.J. Thali. 2008. Whole body post-mortem angiography with a high viscosity contrast agent solution using poly ethylene glycol as contrast agent dissolver. J Forensic Sci 53:465–468. 20. Perrson, A. 2007. Virtual autopsies guide postmortem investigation. Diag Imaging Europe 20–28. 21. Poulsen, K. and J. Simonsen. 2007. Computed tomography as routine in connection with medico-legal autopsies. Forensic Sci Int 171:190–97. 22. Bisset, R. 1998. Magnetic resonance imaging may be alternative to necropsy. BMJ 317:1450. 23. Bisset, R., N. B. Thomas, I. W. Turnbull, and S. Lee. 2002. Postmortem examinations using magnetic resonance imaging: four year review of a working service. BMJ 324:1423–24. 24. Brookes, J. S. and C. Hagmann. 2006. MRI in fetal necropsy. J Magn Reson Imaging 24:1221–28. 25. Brookes, J. S. and M. A. Hall-Craggs. 1997. Postmortem perinatal examination: the role of magnetic resonance imaging. Ultrasound Obstet Gynecol 9:145–47. 26. Brookes, J. A., M. Hall-Craggs, and W. R. Lees. 1999. Magnetic resonance necropsy is offered routinely in university college London hospitals. BMJ 319:56–57. 27. Brookes, J. A., M. A. Hall-Craggs, V. R. Sams, and W. R. Lee. 1996. Non-invasive perinatal necropsy by magnetic resonance imagining. Lancet 348:1139–41. 28. Cohen, M., M. Paley, P. Griffiths, and E. Whitby. 2007. Less invasive autopsy: benefits and limitations of the use of magnetic resonance imaging in the perinatal post-mortem. Pediatr Dev Pathol 1. 29. Griffiths, P. D., M. N. Paley, and E. H. Whitby. 2005. Postmortem MRI as an adjunct to fetal or neonatal autopsy. Lancet 365:1271–73. 30. Hagmann, C. E., N. J. Robertson, V. R. Sams, and J. A. Brookes. 2007. Postmortem magnetic resonance imaging as an adjunct to perinatal autopsy for renal-tract abnormalities. Arch Dis Child Fetal Neonatal Ed 92:F215–18. 31. Roberts, I. S., E. W. Benbow, R. Bisset, et al. 2003. Accuracy of magnetic resonance imaging in determining cause of sudden death in adults: comparison with conventional autopsy. Histopathology 42:424–30. 32. Rutty, G. N. 2005. Are invasive autopsies necessary? Forensic Sci Med Pathol 1:71–73. 33. Rutty, G. N. 2007. Are autopsies necessary? The role of computed tomography as a possible alternative to invasive autopsies. Rechtsmedizin 17:21–28. 34. Rutty, G. N., P. Boyce, C. E. Robinson, A. J. Jeffery, and B. Morgan. 2008. The role of computed tomography in terminal ballistic analysis. Int J Legal Med 122:1–5. 35. Rutty, G. N. and B. Swift. 2004. Accuracy of magnetic resonance imaging in determining cause of sudden death in adults: comparison with conventional autopsy. Histopathology 44:187–89. 36. Swift, B. and G. N. Rutty. 2006. Recent advances in postmortem forensic radiology: CT and MRI applications. In: Forensic Pathology Reviews, vol. 4, ed. M. Tsokos, 355– 404. Totowa, NJ: Humana Press.
© 2009 by Taylor & Francis Group, LLC
477
37. Whitby, E. H., M. N. Paley, M. Cohen, and P. D. Griffiths. 2006. Post-mortem fetal MRI: what do we learn from it? Eur J Radiol 57:250–55. 38. Widjaja, E., E. H. Whitby, M. Cohen, M. N. Paley, and P. D. Griffiths. 2006. Post-mortem MRI of the foetal spine and spinal cord. Clin Radiol 61:679–85. 39. Wilson, C. A., A. K. Bonner, and G. N. Rutty. 2003. Radiological investigation in autopsy practice. In: Essentials of autopsy practice. Recent advances, topics and developments, vol. 2, ed. G. N. Rutty, 129–58. London: Springer Verlag.
D7.4 NORTH AMERICA The medical examiners of the Armed Forces Institute of Pathology in the United States have had a multislice computed tomography (MSCT) scanner since 2005. Some publications from the Armed Forces Institute of Pathology are given in the last years (A.D. Levy). In other institutes in the United States, there is discussion on forensic imaging and virtual autopsies [1–11]. 1. Becker, G. J. 2005. Virtues of virtual autopsy. J Am Coll Radiol 2:376–78. 2. Haglund, W.D., and C. L. Fligner. 1993. Confirmation of human identification using computerized tomography (CT). J Forensic Sci 38:708–12. 3. Harcke, H. T., A. D. Levy, R. M. Abbott, C. T. Mallak, J. M. Getz, and L. Pearse. 2007. Autopsy radiography: digital radiographs (DR) vs multidetector computed tomography (MDCT) in high-velocity gunshot-wound victims. Am J Forensic Med Pathol 28:13–19. 4. Levy, A. D., R. M. Abbott, C. T. Mallak, et al. 2006. Virtual autopsy: preliminary experience in high-velocity gunshot wound victims. Radiology 240:522–28. 5. Levy, A. D., H. T. Harcke, J. M. Getz, et al. 2007. Virtual autopsy: two- and three-dimensional multidetector CT findings in drowning with autopsy comparison. Radiology 243:862–68. 6. Lundberg, G. D. 1998. Low-tech autopsies in the era of high-tech medicine: continued value for quality assurance and patient safety. J Am Med Assoc 280:1273–74. 7. Mitka, M. 2007. CT, MRI scans offer new tools for autopsy. J Am Med Assoc 298:392–93. 8. Oliver, W. R. 1998. Image processing in forensic pathology. Clin Lab Med 18:151–80. 9. Oliver, W. R., A. Boxwala, J. Rosenman, T. Cullip, J. Symon, and G. Wagner. 1997. Three-dimensional visualization and image processing in the evaluation of patterned injuries: the AFIP/UNC experience in the Rodney King case. Am J Forensic Med Pathol 18:1–10. 10. Oliver, W. R., A. S. Chancellor, M. Soltys, et al. 1995. Threedimensional reconstruction or a bullet path: validation by computed radiography. J Forensic Sci 40:321–24. 11. Patriquin, L., A. Kassarjian, M. Barish, et al. 2001. Postmortem whole-body magnetic resonance imaging as an adjunct to autopsy: preliminary clinical experience. J Magn Reson Imaging 13:277–87.
D8
Miscellaneous
CONTENTS D8.1 Paleoradiology .............................................................................................................................................................. 479 D8.1.1 Introduction ...................................................................................................................................................... 479 D8.1.2 General Principles of Mummification .............................................................................................................. 480 D8.1.2.1 What Kinds of Postmortem Decay Processes Need To Be Arrested by the Different Mummification Principles? .................................................................................. 480 D8.1.2.2 But How Can These Processes Be Slowed or Almost Stopped? .................................................... 480 D8.1.3 Egyptian Human Mummies ............................................................................................................................. 480 D8.1.4 Egyptian Animal Mummies ............................................................................................................................. 486 D8.1.5 South American Mummies .............................................................................................................................. 488 D8.1.6 References......................................................................................................................................................... 490 D8.2 Forensic Veterinary Radiology ...................................................................................................................................... 491 D8.2.1 Introduction ...................................................................................................................................................... 491 D8.2.2 Ballistic Trauma ............................................................................................................................................... 491 D8.2.2.1 Case 1............................................................................................................................................... 491 D8.2.2.2 Case 2............................................................................................................................................... 493 D8.2.3 Sharp Trauma .................................................................................................................................................. 493 D8.2.3.1 Case 3............................................................................................................................................... 494 D8.2.3.2 Case 4............................................................................................................................................... 494 D8.2.3.3 Case 5............................................................................................................................................... 496 D8.2.4 Blunt Trauma .................................................................................................................................................... 498 D8.2.4.1 Case 6............................................................................................................................................... 498 D8.2.5 Conclusion ........................................................................................................................................................ 498
D8.1 PALEORADIOLOGY Christian Jackowski, Stephan Bolliger, and Michael J. Thali
D8.1.1 INTRODUCTION Because the goal of paleopathological research is to find an ideal compromise between investigating an object while not destroying it, noninvasive investigation techniques are of inestimable value. Plain radiography was often used to make the contents of wrapped mummies visible without performing dissection while still projecting the 3D information of the mummy onto a 2D x-ray film. Thereby skeletal features could be seen, but the characterization of wrappings, contents, mummification techniques, and so forth was less satisfactory. In 1979 [1], computed tomography (CT) proved to be the most valuable noninvasive investigation tool for mummy research, and it has remained the noninvasive investigation technique of choice for mummies today. CT allows noninvasive insight, while dissection destroys characteristic features of the object and often obviates exposure to subsequent publicity, as after the unrollings and mummy autopsies that have been performed in the last centuries.
CT imaging supplies the 3D structural information of the scanned mummified object. Cross-sections in any desired direction through the mummy are possible; in addition, 3D images and models can be achieved. Virtual endoscopy allows for a journey through unwrapped bodies; hence, detailed information about sex, age, constitution, injuries, possible diseases, and techniques used for mummification can be obtained. Many mummified human remains from various former cultures have undergone CT examination, and the findings have been published [2–7]. In recent years more substantial medium-sized studies extended the knowledge about mummification features and the individual mummies [8–14]. An extensive and detailed description of cross-sectional features of Egyptian mummies can be found, for example, in a recent published textbook by Raven and Taconis [15], and a broad overview on geographic mummification characteristics is given by Arthur C. Aufderheide in his textbook The Scientific Study of Mummies [16]. Since the Institute of Forensic Medicine at the University of Bern owns a CT unit, it became an ideal setting for anthropological or historical radiological investigations. Neither patient schedules nor hygienic concerns can complicate the access to the technology for less clinical disciplines such as anthropology or paleopathology. Thereby close collaborations 479
© 2009 by Taylor & Francis Group, LLC
480
were possible among our forensic and anthropological institutes, museums, and private collections in Switzerland [24]. At the end of this introduction, we would like to address readers who may wonder why we present only CT images or data, and who miss MR images, in section D8.1, “Paleopthology,” keeping in mind that both techniques are implemented within the structure of the Virtopsy Project. Although magnetic resonance imaging (MRI) also increases the soft-tissue resolution postmortem, it fails in imaging the desiccated tissue of mummies [10]. MRI predominantly depends on the hydrogen protons within the tissue H2O, and the dehumidified mummy tissues cannot generate sufficient signal for adequate MRI. Comparable experiences were shown for adipocere bodies when putrefaction gas and calcifying adipocere formation reduced the tissue signal and obviated sufficient MRI [17].
D8.1.2 GENERAL PRINCIPLES OF MUMMIFICATION Every biological material undergoes decay processes after the death of the individual. Mummification characterizes the mechanisms that transform the tissue into a state of decelerated and almost arrested decay. Mummification processes can be summarized as “artificial” when procedures were performed on the corpse with the aim of mummification, or as “natural” when the natural environment of the corpse has resulted in preservation. Mummification of human remains always aimed to preserve the living morphology for a prolonged postmortem interval.
The Virtopsy Approach
tissue preservation of human corpses. Furthermore, exposure to heat or radiation can result in a bactericide effect and thereby slow the putrefaction. This phenomenon can be seen when bodies or body parts that remained in the hot summer sun get quickly mummified. Otherwise, rapid desiccation or dehydration will withdraw the tissue water needed for every biological process in natural decay. Very dry regions on our planet, such as the Atacama Desert in northern Chile, the Sahara Desert in Egypt, or the Gobi Desert in China, therefore show an increased occurrence of natural mummification. The Egyptians employed artificial tissue desiccation techniques most impressively. Knowing the hydrophilic nature of natron, they enveloped the bodies after evisceration with natron for several weeks (more extensively described later). This resulted in a sufficient water transfer from the body tissues to the natron, leading to an artificial mummification by desiccation. Every condition leading to a rapid loss of water can result in mummification. Surprisingly, even honey has a desiccative effect. This is explained by the high sugar concentration in honey and its osmotic effect at the skin barrier as a semipermeable membrane. Thereby tissue water will go into the honey to dilute the highly concentrated sugar that cannot likewise pass the skin. However, it is doubtful whether this has ever been implemented on human mummies. Further mechanisms include the chemical behavior of, for example, heavy metals, resins, lime, or lye, which also have bactericide effects.
D8.1.3 EGYPTIAN HUMAN MUMMIES D8.1.2.1 What Kinds of Postmortem Decay Processes Need To Be Arrested by the Different Mummification Principles? After the death of a living being, an enzymatic process of proteolysis begins immediately. The enzymes that break down first the larger proteins, are set free from the body’s own reservoirs, a process called autolysis (i.e., self-destruction). Later, the bacterial florae from the intestine overgrow the tissues of the corpse. These bacteria also produce proteolytic enzymes that result in further tissue decay, called putrefaction. Without going into detail here (any comprehensive forensic textbook can be consulted), these processes normally result in a complete disintegration of the body tissues.
Mummification in ancient Egypt goes back to the Predynastic Period (5500–3050 b.c.) (Table D8.1.1), although the rarely survived mummies from that time mummified spontaneously. The first hints of artificial mummification attempts are discussed for the Archaic Period (3050–2663 b.c.). However, what we mean today when talking about Egyptian mummies began during the Old Kingdom (2663–2195 b.c.). The
TABLE D8.1.1 Ancient Egypt’s Chronology
D8.1.2.2 But How Can These Processes Be Slowed or Almost Stopped? Various conditions need to be achieved, or sometimes occur naturally, that are able to influence the natural decay processes negatively. The first and most common form is the thermal effect (e.g., heat or cooling), which can result in slowing the proteolysis. The aim of every refrigerator (for personal use in our kitchens at home, or for corpse storage in a pathological or forensic institution) utilizes the decay-slowing effect of low temperatures. Steady temperatures below 0°C, as in permafrost or in glaciers, result in spontaneous
© 2009 by Taylor & Francis Group, LLC
From Aufderheide, A. (2003).
Miscellaneous
combination of evisceration followed by desiccation using natron and use of liquid resin to stop the growth of putrefaction bacteria and prevent rehydration became so effective for body preservation that the embalmed bodies endured for nearly 4,500 years, up until the present [16]. To introduce the subject of the cross-sectional investigation of Egyptian mummies, first a summary is presented of how mummification was carried out in ancient Egypt, and how the process developed. The information given here is grounded on the information gained from the unwrappings of mummies in the past as well as from the modern imaging of the last hundred years. Some original sources of information exist, but the ancient Egyptians were rather reluctant to describe their procedures. This is in contrast to the medical, mathematical, literary, or religious subjects showing a considerable amount of information that has been stored predominantly on papyrus. Many tombs present with decorations depicting preparations that were necessary for burial, but details about exact procedures and the various phases are missing. Written information was found in several papyri dating to the Roman Period, and known as the Ritual of Embalming. Some papyri contain instructions for wrapping, or information on the duration of several stages. Others contain information about the order in which the extremities have to be wrapped, and how amulets are inserted. Other papyri, dating to 200 b.c., give more detailed information on the duration of the various stages of the funerary rites and the position of the incisions for subcutaneous packing. The first detailed description of the practice of mummification was written by Herodotus in the fifth
481
century b.c. Herodotus specifies various mummification procedures and indicates three classes of evisceration and burial, according to price [18,19]. The significance of mummification was preservation of the corpse to allow the soul to recognize it and to reunite with it in the afterlife. Artificial mummification was initially reserved for the pharaoh and his nobles, but then the practice spread. The processes of artificial mummification are generally summarized as follows [20,21]: 1. The body was laid out on a wooden table, and the clothes were removed, followed by purification by washing with a natron solution. This was done in a temporary structure near a body of water such as the Nile River or a canal. Apart from practical issues, purification of the corpse had an important ritual meaning. 2. The embalmers then removed most of the internal organs, knowing that the decomposition starts herein. Thereby they prevented early disintegration of the whole body. Often removal of the brain was performed via the nostrils by perforating the cribriform plate (Figure D8.1.2). Usually only one side of the nose was used, but sometimes this procedure was executed in a rather rough way, leading to disintegration of the adjacent anatomical structures like the conchae, the orbital wall, or the ethmoid cells. Long metallic or wooden hooks were advanced through the nose and the cribriform plate into the cranial cavity. Removal was most
FIGURE D8.1.1 Often only parts of Egyptian mummies remain and can undergo CT examination. Besides hands, legs, or braids of hair, the head is the most frequently surviving mummy part, either totally or partly unwrapped.
© 2009 by Taylor & Francis Group, LLC
482
The Virtopsy Approach
FIGURE D8.1.2 The most common finding in Egyptian mummy heads is a destruction of the cribriform plate and an empty or sometimes refilled skull. This is the way the embalmers usually removed the brain from the skull of the deceased. Sagittal multiplanar reformatting (upper images) as well as volume-rendered 3D models of the skull (lower images) clearly demonstrate the defect of the cribriform plate (arrows) and show no remnants of brain remaining within the skull.
probably effected by rinsing the macerated brain out of the cranium via the created bony defect. Not yet knowing its significant meaning for the individual’s life, the brain was often disposed of at that time. The skull was then rinsed with fluids. The dura and the falx often remained within the skull. The empty skull was then packed with sheets of linen, sand, mud, or resin, or was left empty. An
alternative and distinctively less used method for brain removal was to enlarge an approach through the foramen magnum at the base of the skull (see also the mummy from South America shown in Figure D8.1.20). 3. Removal of the thoracic and abdominal organs was performed by creating a short left abdominal incision. The kidneys, heart, and great vessels
FIGURE D8.1.3 When brain removal was not performed during embalming, the entire base of the skull remained intact, as seen within the sagittal multiplanar reformatting (left) and the volume-rendered 3D model (middle) of the skull base (arrows). Both hemispheres of the brain can then be found shrunken within the dorsal region of the skull (right).
© 2009 by Taylor & Francis Group, LLC
Miscellaneous
483
FIGURE D8.1.4 Sometimes CT scanning reveals small surprises. In this case, a remnant of a wooden tool for brain removal was seized within the sphenoid bone. To rule out the possibility that during unwrapping someone put something into the mummy’s nose, the wooden piece was extracted. Radio carbon dating of the wooden piece revealed an age of approximately 2,200 years (388–196 b.c.). This leads to the conclusion that not only metallic tools such as hooks were used for the procedure of brain removal, as some authors describe.
FIGURE D8.1.5 CT scanning shows pathological skeletal and mummified corpse findings. One female skull showed a variant of the skull sutures with an additional one through the frontal bone called metopic suture or sutura frontalis persistens (arrows). It is a remnant of the so-called frontal suture in infants and children as a dense connective tissue structure dividing the two halves of the frontal bone. It usually disappears at the age of six, when the two halves of the frontal bone have fused together. However, this has no pathological meaning, but can make a relation to other individuals with the same finding more likely as it shows familiar accumulations.
FIGURE D8.1.6 Multiplanar reformatting and volume-rendered 3D model of the cervical spine of the female mummy head as shown in Figure D8.1.5 shows hints of spina bifida, with unclosed vertebral arches of the 4th and 5th cervical vertebra.
© 2009 by Taylor & Francis Group, LLC
484
The Virtopsy Approach
FIGURE D8.1.7 Axial CT image of a mummy head shows a missing frontal sinus on the right side (arrow). This has no pathological meaning.
usually remained within the corpse, as the intestine was known to be the predominant cause of rapid putrefaction. An alternative method of extracting the viscera was by injecting dissolving fluids into the rectum, which were then extruded through the rectum. 4. The empty body and the removed organs were rinsed with water and sometimes also treated with spices, which had a sterilizing effect. The removed organs were divided into four groups—lungs, liver, stomach, and intestines—and these were embalmed separately. They were either placed in special containers (canopic jars) or replaced into the body cavities in the form of packages after embalming. To ensure the possibility of further use of these organs in the afterlife, their safety was entrusted to the Sons of Horus. Each of them took care of one organ: Imsety, human-headed protector
of the liver; Hapy, baboon-headed protector of the lungs; Duamutef, jackal-headed protector of the stomach; and Qebehsenuef, falcon-headed protector of the intestines. The cap of each of the canopic jars depicted its corresponding deity. 5. The most important and probably most lengthy phase was the desiccation or dehydration of the body with natron, a compound of sodium salts such as sodium carbonate, bicarbonate, sulfate, or chloride (Figure D8.1.9). It was usually applied in packages of linen that were stuffed in the body cavities, and the entire body surface was covered with it. Zimmerman [22] carried out experimental artificial mummification on a human body under circumstances adapted to those in ancient Egypt. He describes that 273 kg of Egypt’s natron was not enough for a complete dehydration of an eviscerated human body weighing 70.9 kg, although the weight was reduced to 35.9 kg and several parts of the corpse were preserved very well during the desiccation in 35 days. Herodotus noted that the dehydration took around 40 days (out of a total of 70 days for the embalming). 6. After dehydration, the muscles and subcutaneous fatty tissue almost completely disappear, leaving the body as a bony skeleton covered with a thin, wrinkled skin. The body cavities were then rinsed and packed with filling material in order to remodel the former body shape. Different types of material were used such as mud, sheets, sand, or rolls of linen, resin, and also the visceral packages, when the removed organs were placed back in the body and not stored in canopic jars. The body was then prepared with oils, aromatic resins, unguents, and perfumes. Liquefied resin was spread over the whole body to prevent rehydration. The physical appearance of the deceased was reconstructed
FIGURE D8.1.8 Egyptian “unwrapped” mummy of a female child who is approximately 8 months old and from the Ptolemaic or Roman Period (radio carbon dated a.d. 18–a.d. 134). Remnants of the wrapping remained on the right eye. The mummy was originally covered with gold dust and shows reddish henna coloration of the hair.
© 2009 by Taylor & Francis Group, LLC
Miscellaneous
485
FIGURE D8.1.9 Sagittal multiplanar reformatting of the skull (left) shows an empty cranial cavity, an intact cribriform plate, and granular filling material within the oral cavity (arrow). Some of this material was also found in the nasal cavities (dashed arrow). Analysis of the filling material showed crystalline sodium chloride as its major component. An oblique cranio-caudal view on a maximum intensity projection (MIP) model (right) also demonstrates the filling material.
FIGURE D8.1.10 A very unusual technique for brain removal was applied in this 8-month-old female child. Instead of the well-known transnasal or less widespread atlanto/foramen magnum approach, a hole was made in the skull of the child right behind and above the left ear (arrows). This is not visible at external inspection, as the hair at the cutaneous incision is very tightly fixed to the skull, probably due to the reddish henna coloration. Note that the cribriform plate shows no damage (dotted arrow).
FIGURE D8.1.11 Thoracic and abdominal cavities are empty. No cutaneous incision could be radiologically found. External inspection also failed to demonstrate an incision. The incision is probably somewhere within the skin folds on the thorax (Figure D8.2.8).
© 2009 by Taylor & Francis Group, LLC
486
The Virtopsy Approach
1 2 3 4 5 6 7
?
8 9 10 11 12
FIGURE D8.1.12 The second unusual finding concerns the bony thorax. The 3rd rib on the right side has been removed and placed between 6th and 7th rib on the right side, as seen in these MIP-models. The 6th and 7th ribs are pushed apart. The tip of the 3rd rib is placed between 4th and 5th ribs parasternal. This seems to be intended to protect the thoracic cavity from collapse in supine position (note the right images). This thoracic approach was probably also used for evisceration.
by arranging the hair, inserting artificial eyes, or applying subcutaneous packing material. 7. The last phase was the wrapping of the mummy. Besides keeping the integrity of the corpse, this mainly had a religious meaning. Fine linen was predominantly used for royal mummies, and material of everyday use was used for the deceased of
a lower social position. The quantity of the wrappings and the length and thickness of the bandages varied. Usually, the head and extremities were wrapped separately. Afterward, bandages followed, which mostly showed the known spiral course. A shroud was then placed on the wrapped mummy. Jewelry and funerary amulets can be found on the corpse or between the wrappings. The wrapping and preparation of the entombment are considered to have taken the second half the 70-day time frame reported by Herodotus. This is a general summary of the aspects of mummification in ancient Egypt, and we do not claim it to be all-embracing. The techniques and materials used vary to some extent from one time period to another. The knowledge about differences in the techniques applied to the corpses can be used by well-skilled Egyptologists to relate a mummy to a specific period. Detailed information on that is given in, for example, Aufderheide [20].
D8.1.4 EGYPTIAN ANIMAL MUMMIES
FIGURE D8.1.13 Unwrapped mummy of a crocodile. On the right side, a posterior-anterior view and a sagittal view on an MIP reconstruction of the data demonstrate the skeleton of the hornback skin with its scale pattern.
© 2009 by Taylor & Francis Group, LLC
Four different categories of animal mummies can be distinguished in Egypt’s history. The first group consists of animals that were conserved as food for the deceased. To the second category of animal mummies belong the quite rare cases of animals that accompany the deceased as pets. The third category consists of animals that were seen as corporal manifestations of certain deities. These were allowed to die a natural death and were embalmed afterward. The last group contains those formed by votive offerings that
Miscellaneous
487
FIGURE D8.1.14 Upper cranial bones show multiple fractures (arrows), possibly due to killing of the animal. A broad incision (arrow within the right sagittal multiplanar reformatting) shows the way the embalmers removed the reptile’s brain, as there is no brain remaining within the skull.
were donated by devotees or pilgrims. These animals were bred commercially for this purpose and intentionally killed before embalming. This group included a wide variety of animals, such as dogs, cats, jackals, falcons, hawks, ibises, cows, sheep, crocodiles, snakes, and fish. This early commercialization also resulted in fake mummies that were just wrappings without any remains of animals, or only loose bones, feathers, and tatters bundled up to look like the desired animal. Radiological investigation can easily separate real from unreal and prove the authenticity of any animal bundle (Figure D8.1.17).
Various methods of mummification have been practiced. Some animal mummies (e.g., birds, cats) were shaped to look like human mummies, especially children. Others remained in their shapes also as a bundle. The practice of mummifying animals disappeared when Christianity spread during the 4th century a.d. One of the largest animal mummy collections can be seen in the Egyptian Museum in Cairo. These have been scientifically investigated in the Animal Mummy Project (AMP) [23]. In this study a visual examination of both wrapped and unwrapped mummies, possible zoological classification, x-ray examination, and photographic
FIGURE D8.1.15 Two radioopaque structures were found within the abdomen of the crocodile. Hounsfield units (HU) were 1,500 HU and 1900 HU. The two foreign bodies show the appearance of two stones that probably found their way into the reptile’s intestine during ingestion. This is a normal finding, as alligators swallow stones to decrease buoyancy. Furthermore, a fracture of the right hind leg that occurred either shortly prior to death or postmortem was revealed. Note that there is no callus formation seen at this fracture.
© 2009 by Taylor & Francis Group, LLC
488
The Virtopsy Approach
FIGURE D8.1.16 Although parts of the internal organs of the scanned reptile remained within the mummy, there are some regions that show distinctively less abdominal content. Close to these regions holes within the skin can be detected that go through all layers of the abdominal wall (arrows). A longer cutaneous incision could not be found. This leads to the conclusion that the ancient embalmers may have utilized hooks or comparable tools to eviscerate the reptile.
documentation were performed. Age and cause of death were studied to the extent possible, and the wrappings were described.
D8.1.5 SOUTH AMERICAN MUMMIES Artificial mummification was not only performed by the well-known ancient Egyptian embalmers. Human mummies of at least as good preservation are also found on almost
HAL
HRA
every continent. South and Middle American tribes left a lot of mummies to posterity in the territories of today’s Mexico, Chile, Peru, Ecuador, and Colombia during 1800 b.c. to a.d. 1500. But mummification started far earlier in South America. Roughly 7,000 years ago, or 2,000 years before the Egyptians started artificial mummification, a small Andean tribe known as the Chinchorros began to mummify its dead using elaborate preparatory processes. Although some naturally occurring mummies, which were dried by the hot climate
HRA
HPL
FIGURE D8.1.17 Besides reptiles, thousands of birds, predominantly the ibis and falcon, were mummified in Egypt. Two mummies of birds are seen. The left suffered detachment of the head. The CT examination ruled out any fake mummy in these two cases.
© 2009 by Taylor & Francis Group, LLC
Miscellaneous
489
FIGURE D8.1.18 Mummy from the highlands of the Andes. Flexed extremities are typical for South and Middle American mummies. Note also the deformation of the cranium in the lateral view. Several tribes along the Andes marked their individuals as a member of a certain group by a deformation of the head. This deformation was achieved by placing a textile band around the head during early childhood. This band was then continually tightened during the first two years of life and was usually removed before the third birthday. Note also the postmortem loss of two teeth that remained in the oral cavity (arrow). The bottom images demonstrate a “virtual undressing” by removing the density values of the clothes and soft tissues within the volume-rendered 3D model. Thereby one can have a detailed look at the skeleton of this young male.
conditions of the Atacama Desert in today’s Chile, have been associated with the Chinchorros, archaeologists have discovered many artificially preserved Chinchorro mummies. The oldest were dated back further than 5,000 b.c.
The Chinchorro mummification techniques largely involved preserving the skeleton of a deceased tribe member and “rebuilding” the individual using ritually applied natural materials as well as sculpted clay fixtures and adornments.
FIGURE D8.1.19 Axial cross-section of the abdominal and thoracic cavity. A right abdominal incision (arrow) seems to be the approach through which the abdominal organs have been removed. At the left side of the thorax, a cutaneous incision and displacement of the ribs become obvious. As the thorax is empty, this is the most probable route of thoracic organ extraction.
© 2009 by Taylor & Francis Group, LLC
490
FIGURE D8.1.20 The skull is missing remnants of brain. There is no bony defect within the skull bone but there is a large displacement of the first cervical vertebra. The dorsal space between the arch of the atlas and the base of the skull is distinctly enlarged. This approach for brain removal is described for the embalmers in Egypt as well.
Perhaps most strikingly, the Chinchorros “painted” a large number of mummies using red or black colors or simple mud, and the progression or these coloration techniques has been helpful in tracking the time course of Chinchorro mummification practices. Among the differences between Chinchorro mummification practices and those of ancient Egypt is the broad spectrum of society represented by its mummies. Virtually every Chinchorro was eligible for mummification, including stillborn children. While Egyptian mummification was largely an expensive, status-based privilege reserved for nobles, Chinchorro mummification was for everyone. Anthropologists believe the mummification practice to be part of the Chinchorro’s cultural ancestor worship. The tribe’s mummification practices died out in the first century b.c.
The Virtopsy Approach
Other tribes, such as the Incas, continued their own mummification practices right up until the Spanish invasion of South America in the late 1500s and early 1600s. Surviving examples of these and other tribes’ ritual mortuary practices show that South America should be recognized as the true motherland of artificial human mummification. Mummification started here distinctively earlier, and it lasted much longer compared with the more commonly known Egyptian mummification. South American mummies mostly show flexed extremities and were buried in so-called bundles in some kind of crouch position. Evisceration was not that widespread compared with Egypt’s dynasties, but also was performed. For detailed information on time schedules and tribe characteristics in South American mummification, consult, for example, Aufderheide [20].
D8.1.6 REFERENCES 1. Harwood-Nash D.C. 1979. Computed tomography of ancient Egyptian mummies. J Comput Assist Tomogr 3:768–73. 2. Cesarani F., Martina M.C., Grilletto R., et al. 2004. Facial reconstruction of a wrapped Egyptian mummy using MDCT. Am J Roentgenol 83:755–58. 3. Hughes S., Wright R., Barry M. 2005. Virtual reconstruction and morphological analysis of the cranium of an ancient Egyptian mummy. Australas Phys Eng Sci Med 28:122–27. 4. Motamed M., Alusi G., Campos J., et al. 1998. ENT aspects of the mummification of the head in ancient Egypt: an imaging study. Rev Laryngol Otol Rhinol (Bord) 119:337–39. 5. Baldock C., Hughes S.W., Whittaker D.K., et al. 3D reconstruction of an ancient Egyptian mummy using x-ray computer tomography. J R Soc Med 87:806–08. 6. Pickering R.B., Conces D.J., Jr., Braunstein E.M., et al. 1990. Three-dimensional computed tomography of the mummy Wenuhotep. Am J Phys Anthropol 83:49–55.
FIGURE D8.1.21 The young male suffered a vital fracture of the right femoral neck and a fracture of the 4th lumbar vertebra. This fracture combination indicates a fall from a height, landing on the feet as a possible mechanism. The accident occurred long before the person died, as proved by the distinctive bony transformation at the fractured edges of the femur. This fracture furthermore indicates that the deceased could not walk adequately after the fall until the time of his death.
© 2009 by Taylor & Francis Group, LLC
Miscellaneous
491
7. Vahey T., Brown D. 1984. Comely Wenuhotep: computed tomography of an Egyptian mummy. J Comput Assist Tomogr 8:992–97. 8. Cesarani F., Martina M.C., Ferraris A., et al. 2003. Wholebody three-dimensional multidetector CT of 13 Egyptian human mummies. Am J Roentgenol 180:597–606. 9. Marx M., D’Auria S.H. 1986. CT examination of eleven Egyptian mummies. Radiographics 6:321–30. 10. Notman D.N., Tashjian J., Aufderheide A.C., et al. 1986. Modern imaging and endoscopic biopsy techniques in Egyptian mummies. Am J Roentgenol 146:93–96. 11. Hoffman H., Hudgins P.A. 2002. Head and skull base features of nine Egyptian mummies: evaluation with high-resolution CT and reformation techniques. Am J Roentgenol 178:1367–76. 12. Hoffman H., Torres W.E., Ernst R.D. 2002. Paleoradiology: advanced CT in the evaluation of nine Egyptian mummies. Radiographics 22:377–85. 13. Gaafar H., bdel-Monem M.H., Elsheikh S. 1999. Nasal endoscopy and CT study of Pharaonic and Roman mummies. Acta Otolaryngol 119:257–60. 14. Jansen R.J., Poulus M., Taconis W., et al. 2002. Highresolution spiral computed tomography with multiplanar reformatting, 3D surface- and volume rendering: a non-destructive method to visualize ancient Egyptian mummification techniques. Comput Med Imaging Graph 26:211–16. 15. Raven M.J., Taconis W. 2005. Egyptian Mummies— Radiological Atlas of the Collections in the National Museum of Antiquities in Leiden. Turnhout Belgium: Brepol Publishers. 16. Aufderheide A.C. 2003. The Scientific Study of Mummies. Cambridge: Cambridge University Press. 17. Jackowski C., Thali M., Sonnenschein M., et al. 2005. Adipocere in postmortem imaging using multislice computed tomography (MSCT) and magnetic resonance imaging (MRI). Am J Forensic Med Pathol. 26:360–64. 18. Brier B. 1994. Egyptian Mummies: Unraveling the Secrets of an Ancient Art. New York: Harper. 19. Taylor J.H. 2001. Death and the Afterlife in Ancient Egypt. Chicago: Chicago University Press. 20. Aufderheide A.C. 2003. The Scientific Study of Mummies. Cambridge, UK: Cambridge University Press. 21. Raven M.J., Taconis W. 2005. Egyptian Mummies— Radiological Atlas of the Collections in the National Museum of Antiquities in Leiden. Turnhout Belgium: Brepol Publishers. 22. Zimmerman M.R., Brier B., Wade R.S. 1998. Brief communication: twentieth-century replication of an Egyptian mummy: implications for paleopathology. Am J Phys Anthropol 107:417–20. 23. Ikram S., Iskander N. 2002. Non-Human Mummies. Cairo: American University in Cairo Press. 24. Jackowski, C., Bolliger, S., and Thali, M.J. 2008. Common and unexpected findings in mummies from ancient Egypt and South America as revealed by CT. Radiographics Sept– Oct; 28(5):1477–92.
D8.2 FORENSIC VETERINARY RADIOLOGY Stephan A. Bolliger and Michael J. Thali
D8.2.1 INTRODUCTION Forensic medicine and pathology deals mainly with the examination of harm inflicted upon human beings. However, with
© 2009 by Taylor & Francis Group, LLC
the introduction of new laws governing the treatment of vertebrates in most countries and a heightened public awareness of these issues, the need for expert testimonies on animal cruelty and poaching has arisen. Indeed, this has created an interesting collaboration between veterinary pathologists, who by virtue of their training know the anatomy and pathologic changes of the creatures to be examined, and forensic pathologists, who are trained in seeking and finding clues as to the identity of a possible perpetrator and reconstructing the course of events. This clue finding and crime reconstruction can be undertaken using more or less the same methods as in the examination of human cases: radiology, autopsy, DNA analysis, and toxicology. However, the state of the animals to be examined is often far worse than in human forensic pathology; the creatures are often sent by post in a plastic bag to the veterinary pathology unit. In summer, this leads to advanced putrefaction and all the problems discussed in Chapter D2.1, “Decomposition.” Furthermore, the animals are already partially or even completely dissected before the responsible veterinary pathologist calls for a forensic pathologist. This situation of piecing together the literal “leftovers” is, naturally, highly frustrating for the forensic pathologist, who is expected to present the findings at court later on. The hereby encountered problems can be corrected to a certain degree by the implementation of postmortem radiology, as shown in the selected cases briefly discussed.
D8.2.2 BALLISTIC TRAUMA Gunshot injuries as well as remaining projectiles in the cadaver can be detected by multislice computed tomography (MSCT) in a likewise fashion as in humans. Obviously, certain restrictions with regard to the animal’s size are inevitable: for example, the whole-body scan of a full-sized stock bull is impossible with standard CT machines. However, in smaller animals, the quality of the results is comparable to the quality achieved in human postmortem MSCT scans, as the following poaching cases clearly demonstrate. D8.2.2.1 Case 1 The Natural History Museum in Bern contacted the Institute of Forensic Medicine in order to examine a gorilla skull that had been donated to the museum. The suspicion had arisen that the animal had been killed by poachers. The skull itself presented no suspect lesions (Figure D8.2.1.1). MSCT was performed, which demonstrated multiple small metallic fragments in the bone (Figure D8.2.1.2 and Figure D8.2.1.3). Such foreign particles are typical for booby trap injuries. Booby traps are usually made from a grenade, an artillery shell, or some other explosive, and are rigged to a trip wire or other form of triggering device. They are often used in poaching but are also widespread in combat regions, most notably in guerrilla warfare. As the skull undoubtedly came from Central or West Africa, regions in which both guerrilla warfare and poaching are widespread, we could not determine whether the gorilla had been killed intentionally or not.
492
The Virtopsy Approach
FIGURE D8.2.1.1 Photograph of the gorilla skull. FIGURE D8.2.1.3 MSCT semitransparent reconstruction. Note the radioopaque objects (yellow arrows) in the skull corresponding to the metallic fragments of a booby trap.
FIGURE D8.2.1.2 MSCT 3D reconstruction.
© 2009 by Taylor & Francis Group, LLC
FIGURE D8.2.2.1 Overview of the fur with four entry wounds on the right (yellow arrows) and three exit wounds on the left side (green arrows).
Miscellaneous
493
FIGURE D8.2.2.4 MSCT 3D reconstruction of the lynx skeleton. The arrows indicate the gunshot injuries of the spine and the ricocheted bullet fragments in the right forelimb.
D8.2.2.2 Case 2
FIGURE D8.2.2.2 Skinned animal cadaver with gunshot wounds in the thorax and lower cervical region.
The cadaver of a young lynx, a protected species in Switzerland that was reintroduced a few decades ago under considerable financial and political efforts, was found in a plastic bag in front of a supermarket. The animal was brought to the veterinary pathology unit, which skinned the animal. After skinning, several “holes” in the animal were detected, upon which the forensic pathologist was called, who in turn performed an MSCT in addition to the already performed conventional x-ray examination. A total of three through-and-through wounds and one retained ricocheted bullet were found in the delivered fur and the already autopsied animal cadaver (Figure D8.2.2.1 and Figure D8.2.2.2). Four entry wounds were seen on the right side of the lynx cadaver. The left side presented three exit-wound groups located in the cervical and the loin region as well as the shoulder blade. MSCT proved superior to the conventional radiographs with respect to reconstruction of the bullet course. The through-and-through wounds of the shoulder blade and the vertebral column could be visualized in addition to the badly deformed ricocheted bullet remains in the right forelimb. Judging by the size of the defect in the shoulder blade, the caliber was estimated as being 5–6 mm (Figure D8.2.2.3, Figure D8.2.2.4, and Figure D8.2.2.5).
D8.2.3 SHARP TRAUMA FIGURE D8.2.2.3 Conventional plane radiograph showing bullet fragments in the right forelimb (arrow). Note the loss of 3D information in the plane radiograph.
© 2009 by Taylor & Francis Group, LLC
MSCT is as useful a tool in veterinary forensic pathology as in human forensic pathology in cases of sharp trauma, as the following three cases demonstrate.
494
The Virtopsy Approach
FIGURE D8.2.2.5 MSCT 3D reconstruction. A penetrating gunshot injury of the left shoulder blade (green arrow) and of the spine (yellow arrow) are depicted.
D8.2.3.1 Case 3 A dog was found dead in his kennel. The veterinarian noted a small, round, penetrating defect on the left side of the thorax. The animal was sent by post to the veterinary pathology unit, where the decomposing creature underwent autopsy. There, collapsed lungs but no defect through the chest could be noted. A forensic pathologist was called to reautopsy the dog seven days later. At the right chest side, a fractured rib in the midst of a large defect of the rib cage was noted (Figure D8.2.3.1 and Figure D8.2.3.2). The veterinary pathologist admitted to have repeatedly probed the defect with instruments and fingers. Postmortem MSCT on the dissected, putrefied remains of the animal showed no foreign bodies (Figure D8.2.3.3, Figure D8.2.3.4, and Figure D8.2.3.5). Thus, a shooting of the animal with a disintegrating remaining projectile, the initial hypothesis, could be ruled out. The animal was most likely stabbed by a pole-like structure and died due to the sustained bilateral pneumothoraces. However, due to the putrefactive state and the previous dissection, we could not determine the instrument used. D8.2.3.2 Case 4
FIGURE D8.2.3.1 Autopsy photograph showing the putrefied cadaver in a supine position.
© 2009 by Taylor & Francis Group, LLC
In summer 2005, claims of a serial animal torturer haunted the region of northwestern Switzerland. In this time span, forensic pathologists were routinely called to assess every suspected case of animal killing. A cow was found with a severely injured tail on a paddock. The vet had to amputate the tail, which was sent directly to the forensic pathologist, who examined the amputated body
Miscellaneous
495
FIGURE D8.2.3.2 Autopsy photograph of the right chest side. Note the large, partitioned defect of the right side of the rib cage.
FIGURE D8.2.3.4 MSCT semitransparent reconstruction of the rib cage. Apart from the fractured rib (arrows), no other pathologic findings are seen.
FIGURE D8.2.3.3 MSCT 3D reconstruction of the rib cage. Note the single fractured rib (arrows). No radioopaque foreign bodies are visible.
© 2009 by Taylor & Francis Group, LLC
FIGURE D8.2.3.5 MSCT axial image of the chest. The lungs (yellow arrows) have collapsed and lie in the dorsal regions of the thoracic cavity. Note the gas (green arrow) in the cardiac chamber, and corresponding to putrefaction gas.
496
The Virtopsy Approach
FIGURE D8.2.4.1 Photograph showing the cow tail and a detail of the cut wound (yellow arrow). The green arrow designates the surgical amputation line.
part and performed an MSCT on it. The tail displayed cleancut edges of an injury system consisting of two parallel cuts (Figure D8.2.4.1). MSCT showed that the neither the cartilage nor the vertebrae of the tail were injured but that the tail was amputated to almost 50% (Figure D8.2.4.2). The wound morphology gave rise to the hypothesis of the tail being cut deliberately by a perpetrator.
D8.2.3.3 Case 5 Another case of a cow being cut occurred near the aforementioned case. Here again, an injured cow was found on a paddock. This time, a deep gash was detected on the udder. The vet was forced to euthanize the suffering animal. The udder was examined by a forensic pathologist, who also performed an MSCT. A deep cut reached from the cranial to the caudal end of the udder, the depth increasing cranially to caudally with a maximal depth of 7 cm (Figure D8.2.5.1). MSCT displayed no foreign bodies within the udder. Angiography with MSCT and an injected contrast agent showed no lesion of major blood vessels (Figure D8.2.5.2). Due to the wound morphology, the cut was assumed to have been inflicted with a knife.
© 2009 by Taylor & Francis Group, LLC
FIGURE D8.2.4.2 MSCT 3D reconstruction showing the surface (top), a semitransparent view (middle), and an osseous reconstruction (bottom). The green arrow designates the surgical lesion, and the yellow arrow designates the cut wound. Note that the bone beneath the cut (yellow circle) is undamaged.
Miscellaneous
497
Caudal
Cranial FIGURE D8.2.5.1 Photograph of the udder seen from below. Note the deep cut reaching from the caudal to the cranial portion. The teats are unharmed.
FIGURE D8.2.5.2 MSCT after injection of a hydrophilic contrast agent into the large blood vessels. The vessels are unharmed. No foreign bodies are visible.
© 2009 by Taylor & Francis Group, LLC
498
The Virtopsy Approach
FIGURE D8.2.6.1 MSCT 3D reconstruction of the skin and the skeleton.
Whether this was due to an animal torturer or for insurance reasons could not be discerned.
D8.2.4 BLUNT TRAUMA D8.2.4.1 Case 6 The Natural History Museum received a dead badger in a plastic bag. The question arose of whether the animal had been poisoned or had died of some other violent death. The cadaver showed no visible injuries to the fur. We refrained from shaving the animal, as it was intended for display later on. MSCT displayed a traumatic disruption of the spinal column at the level of the lumbar vertebrae 3–4. An accumulation of blood was seen in the abdominal cavity (Figure D8.2.6.1, Figure D8.2.6.2, and Figure D8.2.6.3). Due to these images, we concluded that the animal had suffered a blunt trauma to the spine, possibly due to a vehicle accident, and had died due to internal hemorrhage.
D8.2.5 CONCLUSION FIGURE D8.2.6.2 MSCT 3D reconstruction of the lumbar spine. Note the dislocation of the fourth lumbar vertebrate from the third (arrow).
© 2009 by Taylor & Francis Group, LLC
Postmortem imaging of animals is as useful as in human corpses with regard to the detection of signs of ballistic, sharp, or blunt trauma.
Miscellaneous
499
FIGURE D8.2.6.3 MSCT axial image of the abdomen. Note the fluid (arrows), in this case blood, surrounding the intestines. The origin of the hemorrhage is not visible.
© 2009 by Taylor & Francis Group, LLC
Acknowledgments CONTENTS American Journal of Neuroradiology ...................................................................................................................................... 475 American Roentgen Ray Society .............................................................................................................................................. 475 Blackwell Publishing ................................................................................................................................................................ 475 Elsevier Publishers .................................................................................................................................................................... 476 RSNA Publications ................................................................................................................................................................... 477 Springer Science ....................................................................................................................................................................... 477 John Wiley & Sons ................................................................................................................................................................... 477 Wolters Kluwer—Lippincott Williams & Wilkins ................................................................................................................. 477
The editors and authors gratefully thank the following publishers for granting permission to reuse the material from the following articles.
AMERICAN JOURNAL OF NEURORADIOLOGY Yen K, Weis J, Kreis R, Aghayev E, Jackowski C, Thali M, et al. Line-scan diffusion tensor imaging of the posttraumatic brain stem: changes with neuropathologic correlation. Am J Neurorad 2006 Jan;27(1):70–73.
AMERICAN ROENTGEN RAY SOCIETY 44211 Statestone Court Leesburg Virginia 20176-5109 Aghayev E, Sonnenschein M, Jackowski C, Thali M, Buck U, Yen K, et al. Postmortem radiology of fatal hemorrhage: measurements of cross-sectional areas of major blood vessels and volumes of aorta and spleen on MDCT and volumes of heart chambers on MRI. Am J Roentgenol 2006 Jul;187(1):209–15. Grabherr S, Djonov V, Friess A, Thali MJ, Ranner G, Vock P, Dirnhofer R. Postmortem angiography after vascular perfusion with diesel oil and a lipophilic contrast agent. Am J Roentgenol 2006 Nov;187(5):W515–23. Grabherr S, Djonov V, Yen K, Thali MJ, Dirnhofer R. Postmortem angiography: review of former and current methods. Am J Roentgenol 2007 Mar;188(3):832–38. Review.
BLACKWELL PUBLISHING 9600 Garsington Road Oxford, OX4 2DQ, UK Hochmeister MN, Budowle B, Sparkes R, Rudin O, Gehrig C, Thali M, et al. Validation studies of an immunochromatographic 1-step test for the forensic identification of human blood. J Forensic Sci 1999 May;44(3):597–602. Hochmeister MN, Budowle B, Rudin O, Gehrig C, Borer U, Thali M, et al. Evaluation of prostate-specific antigen (PSA) membrane test assays for the forensic identification of seminal fluid. J Forensic Sci 1999 Sep;44(5):1057–60.
Thali MJ, Yen K, Plattner T, Schweitzer W, Vock P, Ozdoba C, et al. Charred body: virtual autopsy with multi-slice computed tomography and magnetic resonance imaging. J Forensic Sci 2002 Nov;47(6):1326–31. Thali MJ, Schwab CM, Tairi K, Dirnhofer R, Vock P. Forensic radiology with cross-section modalities: spiral CT evaluation of a knife wound to the aorta. J Forensic Sci 2002 Sep;47(5):1041–45. Thali MJ, Yen K, Schweitzer W, Vock P, Boesch C, Ozdoba C, et al. Virtopsy, a new imaging horizon in forensic pathology: virtual autopsy by postmortem multislice computed tomography (MSCT) and magnetic resonance imaging (MRI)—a feasibility study. J Forensic Sci 2003 Mar;48(2):386–403. Thali MJ, Taubenreuther U, Karolczak M, Braun M, Brueschweiler W, Kalender WA, et al. Forensic microradiology: micro-computed tomography (Micro-CT) and analysis of patterned injuries inside of bone. J Forensic Sci 2003 Nov;48(6):1336–42. Plattner T, Thali MJ, Yen K, Sonnenschein M, Stoupis C, Vock P, et al. Virtopsy-postmortem multislice computed tomography (MSCT) and magnetic resonance imaging (MRI) in a fatal scuba diving incident. J Forensic Sci 2003 Nov; 48(6):1347–55. Thali MJ, Braun M, Wirth J, Vock P, Dirnhofer R. 3D surface and body documentation in forensic medicine: 3-D/CAD photogrammetry merged with 3D radiological scanning. J Forensic Sci 2003 Nov;48(6):1356–65. Yen K, Vock P, Tiefenthaler B, Ranner G, Scheurer E, Thali MJ, et al. Virtopsy: forensic traumatology of the subcutaneous fatty tissue; multislice computed tomography (MSCT) and magnetic resonance imaging (MRI) as diagnostic tools. J Forensic Sci 2004 Jul;49(4):799–806. Aghayev E, Thali M, Jackowski C, Sonnenschein M, Yen K, Vock P, et al. Virtopsy—fatal motor vehicle accident with head injury. J Forensic Sci 2004 Jul;49(4):809–13. Jackowski C, Thali M, Sonnenschein M, Aghayev E, Yen K, Dirnhofer R, et al. Visualization and quantification of air embolism structure by processing postmortem MSCT data. J Forensic Sci 2004 Nov;49(6):1339–42. Thali MJ, Braun M, Buck U, Aghayev E, Jackowski C, Vock P, Sonnenschein M, et al. VIRTOPSY—scientific documentation, reconstruction and animation in forensic: individual and real 3D data based geo-metric approach including optical body/object surface and radiological CT/MRI scanning. J Forensic Sci 2005 Mar;50(2):428–42. 501
© 2009 by Taylor & Francis Group, LLC
502
Bolliger S, Thali M, Jackowski C, Aghayev E, Dirnhofer R, Sonnenschein M. Postmortem non-invasive virtual autopsy: death by hanging in a car. J Forensic Sci 2005 Mar;50(2):455–60 Jackowski C, Sonnenschein M, Thali MJ, Aghayev E, von Allmen G, Yen K, et al. Virtopsy: postmortem minimally invasive angiography using cross section techniques—implementation and preliminary results. J Forensic Sci 2005 Sep;50(5):1175–86. Thali MJ, Markwalder T, Jackowski C, Sonnenschein M, Dirnhofer R. Dental CT imaging as a screening tool for dental profiling: advantages and limitations. J Forensic Sci 2006 Jan;51(1):113–19. Jackowski C, Bolliger S, Aghayev E, Christe A, Kilchoer T, Aebi B, et al. Reduction of postmortem angiography-induced tissue edema by using polyethylene glycol as a contrast agent dissolver. J Forensic Sci 2006 Sep;51(5):1134–37.
ELSEVIER PUBLISHERS P.O. Box 211, NL-100AE Amsterdam, The Netherlands Oesterhelweg L, Ross S, Spendlove D, Schoen CA, Christe A, Thali MJ, et al. Virtopsy: fatal stab wounds to the skull—the relevance of ante-mortem and post-mortem radiological data in case reconstructions. Leg Med (Tokyo) 2007 Jun;9(6):314–17. Thali MJ, Braun M, Bruschweiler W, Dirnhofer R. Matching tire tracks on the head using forensic photogrammetry. Forensic Sci Int 2000 Sep11;113(1–3):281–87. Thali MJ, Kneubuehl BP, Dirnhofer R, Zollinger U. Body models in forensic ballistics: reconstruction of a gunshot injury to the chest by bullet fragmentation after shooting through a finger. Forensic Sci Int 2001 Nov 15;123(1):54–57. Thali MJ, Kneubuehl BP, Zollinger U, Dirnhofer R. The “skin-skullbrain model”: a new instrument for the study of gunshot effects. Forensic Sci Int 2002 Feb 18;125(2–3):178–89. Thali MJ, Kneubuehl BP, Zollinger U, Dirnhofer R. A study of the morphology of gunshot entrance wounds, in connection with their dynamic creation, utilizing the “skin-skull-brain model.” Forensic Sci Int 2002 Feb 18;125(2–3):190–94. Thali MJ, Kneubuehl BP, Dirnhofer R. A “skin-skull-brain model” for the biomechanical reconstruction of blunt forces to the human head. Forensic Sci Int 2002 Feb 18;125(2–3):195–200. Thali MJ, Kneubuehl BP, Dirnhofer R, Zollinger U. The dynamic development of the muzzle imprint by contact gunshot: high-speed documentation utilizing the “skin-skull-brain model.” Forensic Sci Int 2002 Jul 17;127(3):168–73. Thali MJ, Braun M, Brueschweiler W, Dirnhofer R. “Morphological imprint”: determination of the injury-causing weapon from the wound morphology using forensic 3D/CAD-supported photogrammetry. Forensic Sci Int 2003 Apr 8;132(3):177–81. Thali MJ, Braun M, Markwalder TH, Brueschweiler W, Zollinger U, Malik NJ, et al. Bite mark documentation and analysis: the forensic 3D/CAD supported photogrammetry approach. Forensic Sci Int 2003 Aug 12;135(2):115–21. Kneubuehl BP, Thali MJ. The evaluation of a synthetic long bone structure as a substitute for human tissue in gunshot experiments. Forensic Sci Int 2003 Dec 17;138(1–3):44–49. Plattner T, Kneubuehl B, Thali M, Zollinger U. Gunshot residue patterns on skin in angled contact and near contact gunshot wounds. Forensic Sci Int 2003 Dec 17;138(1–3):68–74.
© 2009 by Taylor & Francis Group, LLC
The Virtopsy Approach
Thali MJ, Yen K, Vock P, Ozdoba C, Kneubuehl BP, Sonnenschein M, et al. Image-guided virtual autopsy findings of gunshot victims performed with multi-slice computed tomography and magnetic resonance imaging and subsequent correlation between radiology and autopsy findings. Forensic Sci Int 2003 Dec 17;138(1–3):8–16. Thali MJ, Yen K, Schweitzer W, Vock P, Ozdoba C, Dirnhofer R. Into the decomposed body-forensic digital autopsy using multislice-computed tomography. Forensic Sci Int 2003 Jul 8;134(2–3):109–14. Brüschweiler W, Braun M, Dirnhofer R, Thali MJ. Analysis of patterned injuries and injury-causing instruments with forensic 3D/CAD supported photogrammetry (FPHG): an instruction manual for the documentation process. Forensic Sci Int 2003 Mar 27;132(2):130–38. Thali MJ, Kneubuehl BP, Zollinger U, Dirnhofer R. A high-speed study of the dynamic bullet-body interactions produced by grazing gunshots with full metal jacketed and lead projectiles. Forensic Sci Int 2003 Mar 27;132(2):93–98. Thali MJ, Braun M, Dirnhofer R. Optical 3D surface digitizing in forensic medicine: 3D documentation of skin and bone injuries. Forensic Sci Int 2003 Nov 26;137(2–3):203–08. Thali MJ, Dirnhofer R. Forensic radiology in German-speaking area. Forensic Sci Int 2004 Sep 10;144(2–3):233–42. Jackowski C, Schweitzer W, Thali M, Yen K, Aghayev E, Sonnenschein M, et al. Virtopsy: postmortem imaging of the human heart in situ using MSCT and MRI. Forensic Sci Int 2005 Apr 20;149(1):11–23. Aghayev E, Thali MJ, Sonnenschein M, Hurlimann J, Jackowski C, Kilchoer T, et al. Fatal steamer accident; blunt force injuries and drowning in post-mortem MSCT and MRI. Forensic Sci Int 2005 Aug 11;152(1):65–71. Jackowski C, Dirnhofer S, Thali M, Aghayev E, Dirnhofer R, Sonnenschein M. Postmortem diagnostics using MSCT and MRI of a lethal streptococcus group A infection at infancy: a case report. Forensic Sci Int 2005 Jul 16;151(2–3):157–63. Aghayev E, Yen K, Sonnenschein M, Jackowski C, Thali M, Vock P, et al. Pneumomediastinum and soft tissue emphysema of the neck in postmortem CT and MRI; a new vital sign in hanging? Forensic Sci Int 2005 Oct 29;153(2–3):181–88. Buck U, Albertini N, Naether S, Thali MJ. 3D documentation of footwear impressions and tyre tracks in snow with high resolution optical surface scanning. Forensic Sci Int 2007 Sep13;171(2–3):157–64. Thali MJ, Kneubuehl BP, Bolliger SA, Christe A, Koenigsdorfer U, Ozdoba C, et al. Forensic veterinary radiology: Ballisticradiological 3D computertomographic reconstruction of an illegal lynx shooting in Switzerland. Forensic Sci Int 2007 Aug 24;171(1) :63–66. Pfaeffli M, Vock P, Dirnhofer R, Braun M, Bolliger SA, Thali MJ. Post-mortem radiological CT identification based on classical ante-mortem X-ray examinations. Forensic Sci Int 2007 Sep 13;171(2–3):111–17. Verhoff MA, Ramsthaler F, Krähahn J, Deml U, Gille RJ, Grabherr S, et al. Digital forensic osteology—possibilities in cooperation with the Virtopsy (R) project. Forensic Sci Int 2007 Apr 20; [Epub ahead of print]. Yen K, Lovblad KO, Scheurer E, Ozdoba C, Thali MJ, Aghayev E, et al. Post-mortem forensic neuroimaging: Correlation of MSCT and MRI findings with autopsy results. Forensic Sci Int 2007 Feb 27. Buck U, Naether S, Braun M, Bolliger S, Friederich H, Jackowski C, et al. Application of 3D documentation and geometric reconstruction methods in traffic accident analysis: with
Acknowledgments
high resolution surface scanning, radiological MSCT/MRI scanning and real data based animation. Forensic Sci Int 2007 Jul 20;170(1):20–28. Sidler M, Jackowski C, Dirnhofer R, Vock P, Thali M. Use of multislice computed tomography in disaster victim identification–advantages and limitations. Forensic Sci Int 2007 Jul 4;169(2–3):118–28. Aghayev E, Thali MJ, Sonnenschein M, Jackowski C, Dirnhofer R, Vock P. Post-mortem tissue sampling using computed tomography guidance. Forensic Sci Int 2007 Mar 2;166(2–3):199–203.
RSNA PUBLICATIONS 820 Jorie Boulevard Oak Brook, Illinois 60523 Dirnhofer R, Jackowski C, Vock P, Potter K, Thali MJ. VIRTOPSY: minimally invasive, imaging-guided virtual autopsy. Radiographics 2006 Sep–Oct;26(5):1305–33. Review.
503
Thali MJ, Ross S, Oesterhelweg L, Grabherr S, Buck U, Naether S, et al. Virtopsy working on the future of forensic medicine. Rechtsmedizin 2007;17:7–12. Verhoff MA, Ramsthaler F, Krähahn J, Gille RJ, Kage P, Kage S, et al. Digitale forensische Osteologie. Rechtsmedizin 2007;17:29–34.
JOHN WILEY & SONS 111 River Street Hoboken, New Jersey 07030-5774 Yen K, Thali MJ, Aghayev E, Jackowski C, Schweitzer W, Boesch C, et al. Strangulation signs: initial correlation of MRI, MSCT, and forensic neck findings. J Magn Reson Imaging 2005 Oct;22(4):501–10.
WOLTERS KLUWER—LIPPINCOTT WILLIAMS & WILKINS 351 West Camden Street Baltimore, Maryland 21201
SPRINGER SCIENCE Postfach 140201, D-14302 Berlin, Germany Aghayev E, Yen K, Sonnenschein M, Ozdoba C, Thali M, Jackowski C, et al. Virtopsy post-mortem multi-slice computed tomography (MSCT) and magnetic resonance imaging (MRI) demonstrating descending tonsillar herniation: comparison to clinical studies. Neuroradiology 2004 Jul;46(7):559–64. Thali MJ, Dirnhofer R, Becker R, Oliver W, Potter K. Is “virtual histology” the next step after the “virtual autopsy”? Magnetic resonance microscopy in forensic medicine. Magn Reson Imaging 2004 Oct;22(8):1131–38. Yen K, Sonnenschein M, Thali MJ, Ozdoba C, Weis J, Zwygart K, et al. Postmortem multislice computed tomography and magnetic resonance imaging of odontoid fractures, atlantoaxial distractions and ascending medullary edema. Int J Legal Med 2005 May;119(3):129–36. Jackowski C, Thali M, Aghayev E, Yen K, Sonnenschein M, Zwygart K, et al. Postmortem imaging of blood and its characteristics using MSCT and MRI. Int J Legal Med 2006 Jul;120(4):233–40. Jackowski C, Aghayev E, Sonnenschein M, Dirnhofer R, Thali MJ. Maximum intensity projection of cranial computed tomography data for dental identification. Int J Legal Med 2006 May;120(3):165–67. Hillewig E, Aghayev E, Jackowski C, Christe A, Plattner T, Thali MJ. Gas embolism following intraosseous medication application proven by post-mortem multislice computed tomography and autopsy. Resuscitation 2007 Jan;72(1):149–53. Yen K, Vock P, Christe A, Scheurer E, Plattner T, Schön C, et al. Clinical forensic radiology in strangulation victims: forensic expertise based on magnetic resonance imaging (MRI) findings. Int J Legal Med 2007 Mar;121(2):115–23. Thali MJ, Jackowski C, Oesterhelweg L, Ross SG, Dirnhofer R. VIRTOPSY—the Swiss virtual autopsy approach. Leg Med (Tokyo) 2007 Mar;9(2):100–04. Virtopsy Team, Oesterhelweg L. Atmosphere of departure in forensic medicine? Rechtsmedizin 2007;17:40–43.
© 2009 by Taylor & Francis Group, LLC
Jackowski C, Thali MJ, Buck U, Aghayev E, Sonnenschein M, Yen K, et al. Noninvasive estimation of organ weights by postmortem magnetic resonance imaging and multislice computed tomography. Invest Radiol 2006 Jul;41(7):572–78. Jackowski C, Lussi A, Classens M, Kilchoer T, Bolliger S, Aghayev E, et al. Extended CT scale overcomes restoration caused streak artifacts for dental identification in CT—3D color encoded automatic discrimination of dental restorations. J Comput Assist Tomogr 2006 May–Jun;30(3):510–13. Jackowski C, Sonnenschein M, Thali MJ, Aghayev E, Yen K, Dirnhofer R, et al. Intrahepatic gas at postmortem computed tomography: forensic experience as a potential guide for in vivo trauma imaging. J Trauma 2007 Apr;62(4):979–88. Thali MJ, Kneubuehl BP, Vock P, Allmen G, Dirnhofer R. Highspeed documented experimental gunshot to a skull-brain model and radiologic virtual autopsy. Am J Forensic Med Pathol 2002 Sep;23(3):223–28. Yen K, Thali MJ, Kneubuehl BP, Peschel O, Zollinger U, Dirnhofer R. Blood-spatter patterns: hands hold clues for the forensic reconstruction of the sequence of events. Am J Forensic Med Pathol 2003 Jun;24(2):132–40. Thali MJ, Schweitzer W, Yen K, Vock P, Ozdoba C, Spielvogel E, et al. New horizons in forensic radiology: the 60-second digital autopsy-full-body examination of a gunshot victim by multislice computed tomography. Am J Forensic Med Pathol 2003 Mar;24(1):22–27. Jackowski C, Thali M, Sonnenschein M, Aghayev E, Yen K, Dirnhofer R. Adipocere in postmortem imaging using multislice computed tomography (MSCT) and magnetic resonance imaging (MRI). Am J Forensic Med Pathol 2005 Dec;26(4):360–64. Aghayev E, Jackowski C, Sonnenschein M, Thali M,Yen K, Dirnhofer R. Virtopsy hemorrhage of the posterior cricoarytenoid muscle by blunt force to the neck in postmortem multislice computed tomography and magnetic resonance imaging. Am J Forensic Med Pathol 2006 Mar;27(1):25–29. Bolliger SA, Thali MJ, Aghayev E, Jackowski C, Vock P, Dirnhofer R, et al. Postmortem noninvasive virtual autopsy: extrapleural hemorrhage after blunt thoracic trauma. Am J Forensic Med Pathol 2007 Mar;28(1):44–47.