Automatic detection of hemorrhagic pericardial effusion on PMCT using deep learning - a feasibility study.

Post mortem computed tomography (PMCT) can be used as a triage tool to better identify cases with a possibly non-natural cause of death, especially when high caseloads make it impossible to perform autopsies on all cases. Substantial data can be generated by modern medical scanners, especially in a forensic setting where the entire body is documented at high resolution. A solution for the resulting issues could be the use of deep learning techniques for automatic analysis of radiological images. In this article, we wanted to test the feasibility of such methods for forensic imaging by hypothesizing that deep learning methods can detect and segment a hemopericardium in PMCT. For deep learning image analysis software, we used the ViDi Suite 2.0. We retrospectively selected 28 cases with, and 24 cases without, hemopericardium. Based on these data, we trained two separate deep learning networks. The first one classified images into hemopericardium/not hemopericardium, and the second one segmented the blood content. We randomly selected 50% of the data for training and 50% for validation. This process was repeated 20 times. The best performing classification network classified all cases of hemopericardium from the validation images correctly with only a few false positives. The best performing segmentation network would tend to underestimate the amount of blood in the pericardium, which is the case for most networks. This is the first study that shows that deep learning has potential for automated image analysis of radiological images in forensic medicine.

3D mug shot-3D head models from photogrammetry for forensic identification.

No human face is like another, not even in monozygotic twins, which makes the face one of the most individualizing characteristic. It is for this reason that the human face is commonly used for identification purposes and police officers take portrait photographs of arrested persons, so-called mug shots. The disadvantage of these 2D mug shots is that the perspective, in which they are taken (usually frontal and lateral-right, left or both), cannot be changed after acquisition, thus limiting a potential comparison between a mug shot and surveillance footage or other visual recordings. Documenting a face in 3D would reduce this problem as it allows adjusting the perspective of the face for image comparisons depending on the needs of the investigator. We have developed a 3D mug shot system containing 26 digital single-lens reflex cameras arranged semi-circularly in a 200° arc with a 1.46 m radius around a height-adjustable chair. We generated photogrammetric models of a test person's face captured by the mug shot system using three different focal lengths settings as well as 3D models of the same face with GOM Atos Triple Scan and Artec Space Spider. The 3D models were then analysed regarding the visibility of detailed morphological features in different regions of the face compared to 2D mug shots. Our results showed that our 3D mug shot system with its photogrammetric documentation generates 3D models with comparable surface quality to Artec-generated models, or even better quality, compared to GOM-generated models. The results of the morphological assessment were affected by the focal length and availability of texture information. In conclusion, the 3D mug shot system is a fast and efficient tool to generate 3D models of the face and may be used in addition to 2D photographs for the purpose of visual forensic identification based on images.

Postmortem imaging of perimortem skeletal trauma.

Various imaging modalities, including conventional radiography, computed tomography, magnetic resonance, and surface scanning have been applied in the examination of skeletal injuries in the forensic context. Although still not a substitute for a full medico-legal autopsy or the examination of skeletal remains, imaging is now increasingly used as a complementary tool in the postmortem analysis of perimortem skeletal trauma. Facilitated by the progress in general computational capacity, multimodal imaging has been proposed for comprehensive forensic documentation. A major advantage of these imaging approaches is that stored digital or physical 3D models of skeletal injuries can be reviewed at any time by various experts as well as be presented in court as evidence to clarify potentially complex medical and forensic aspects of the case. Due to constant technical progress in imaging techniques and software, continuous education, training, and sharing of expertise among engineers, computer scientists, and forensic experts, including forensic pathologists, anthropologists, and radiologists needs to be warranted to maintain high-quality expertise in the detection and interpretation of traumatic injuries on postmortem imaging. The technical developments and ever-improving user-friendliness of 3D imaging and modeling techniques present an atttactive alternative to traditional forensic approaches, but as long as the techniques have not been sufficiently tested and validated for forensic trauma analysis, and best practice manuals for forensic practice are lacking for both the technical procedures and method selection, the use of imaging techniques needs to be reevaluated on a case-by-case basis. In addition, ethical, legal, and financial aspects of the use of imaging and 3D modeling for forensic purposes need to be well understood by all parties in legal proceedings.

Simulation of mirror surfaces for virtual estimation of visibility lines for 3D motor vehicle collision reconstruction.

3D reconstructions of motor vehicle collisions are used to identify the causes of these events and to identify potential violations of traffic regulations. Thus far, the reconstruction of mirrors has been a problem since they are often based on approximations or inaccurate data. Our aim with this paper was to confirm that structured light scans of a mirror improve the accuracy of simulating the field of view of mirrors. We analyzed the performances of virtual mirror surfaces based on structured light scans using real mirror surfaces and their reflections as references. We used an ATOS GOM III scanner to scan the mirrors and processed the 3D data using Geomagic Wrap. For scene reconstruction and to generate virtual images, we used 3ds Max. We compared the simulated virtual images and photographs of real scenes using Adobe Photoshop. Our results showed that we achieved clear and even mirror results and that the mirrors behaved as expected. The greatest measured deviation between an original photo and the corresponding virtual image was 20 pixels in the transverse direction for an image width of 4256 pixels. We discussed the influences of data processing and alignment of the 3D models on the results. The study was limited to a distance of 1.6m, and the method was not able to simulate an interior mirror. In conclusion, structured light scans of mirror surfaces can be used to simulate virtual mirror surfaces with regard to 3D motor vehicle collision reconstruction.

Multi-camera system for 3D forensic documentation.

Three-dimensional (3D) surface documentation is well established in forensic documentation. The most common systems include laser scanners and surface scanners with optical 3D cameras. An additional documentation tool is photogrammetry. This article introduces the botscan© (botspot GmbH, Berlin, Germany) multi-camera system for the forensic markerless photogrammetric whole body 3D surface documentation of living persons in standing posture. We used the botscan© multi-camera system to document a person in 360°. The system has a modular design and works with 64 digital single-lens reflex (DSLR) cameras. The cameras were evenly distributed in a circular chamber. We generated 3D models from the photographs using the PhotoScan© (Agisoft LLC, St. Petersburg, Russia) software. Our results revealed that the botscan© and PhotoScan© produced 360° 3D models with detailed textures. The 3D models had very accurate geometries and could be scaled to full size with the help of scale bars. In conclusion, this multi-camera system provided a rapid and simple method for documenting the whole body of a person to generate 3D data with Photoscan©.