Juergen Fornaro, Abstract Co-Author: Nothing to Disclose

Daniel Eisenhart, Abstract Co-Author: Nothing to Disclose

Nicola Susanne Glaser-Gallion MD, Abstract Co-Author: Nothing to Disclose

Florian Lorenz Glaser-Gallion MD, Presenter: Nothing to Disclose

Simon Wildermuth MD, Abstract Co-Author: Nothing to Disclose

Sebastian Leschka MD, Abstract Co-Author: Nothing to Disclose

Virtual autopsy is a recent method in the field of forensic medicine for the non-invasive examination of bodies after death, using cross-sectional radiological imaging techniques to document pathologies inside the body, however, lacks visualization of the body surface and skin colour. The purpose of this study was to develop and evaluate a visualization method for the fused and complementary display of features in cross-sectional imaging datasets and of the body surface available from image textures taken with digital cameras.

The method was evaluated with four fresh chicken specimens filled with either ballistic gelatine or minced beef to simulate real body compositions. At two chicken each four shots at medium range with a small caliber rifle or four stab wounds with a jackknife were performed. We used a 64-row multidetector computed tomography system (MDCT) and a non-calibrated 12-megapixel digital camera for the data acquisition. The main postprocessing steps included the semiautomatic estimation of the camera parameters, the extraction of the body surface from the MDCT dataset, the registration of image textures onto the extracted surface and the fused rendering of the volumetric data and surface model. Postprocessing times and the mean error of feature point registration were measured.  

All postprocessing steps could be successfully performed. The mean postprocessing time was 4.3min (range 4-5min) and the mean error of feature point registration was 1.8mm (range 0.8-4mm).

The proposed visualization method is time-efficient and accurate.

The proposed method might improve the interpretation of forensic CT by merging real-colour skin surface images with the CT dataset.

Fornaro, J, Eisenhart, D, Glaser-Gallion, N, Glaser-Gallion, F, Wildermuth, S, Leschka, S, Postmortem Analysis of Injury Patterns with Computed Tomography, Enhanced by Appearance Information: A  Pilot Study in Fresh Chicken Specimens.  Radiological Society of North America 2011 Scientific Assembly and Annual Meeting, November 26 - December 2, 2011 ,Chicago IL. http://archive.rsna.org/2011/11016050.html