Is post-mortem CT of the dentition adequate for correct forensic identification?: comparison of dental computed tomograpy and visual dental record INTERNATIONAL JOURNAL OF LEGAL MEDICINE Kirchhoff, S., Fischer, F., LINDEMAIER, G., Herzog, P., Kirchhoff, C., Becker, C., Bark, J., Reiser, M. F., Eisenmenger, W. 2008; 122 (6): 471-479


The gold standard for identification of the dead is the visual dental record. In this context, several authors emphasize computed tomography (CT) as valuable supportive tool for forensic medicine. However, studies focusing on diagnostic accuracy of post-mortem computed tomography (PMCT) are still missing. Therefore, the aim of this study was to compare diagnostic accuracy of the visual dental record and post-mortem computed tomography (PMCT) of the dentition for identification of the dead. Ten whole skulls were included into the study. The entire dentition of each skull was first examined with the visual dental record as a gold standard and second using dental PMCT scans, performed on a 64-multi-detector CT (MDCT). 3D reformations, multi-planar reformations (MPR), and CT-orthopantomography (OPG) were performed in the post-processing. All examinations were analyzed by three independent investigators regarding the criteria for identification of the dead, e.g., in case of disaster. PMCT for the dental identification of the dead was difficult to perform and time consuming. Due to dental overlays and corresponding artifacts, the definite periphery of the dental fillings/inlays was not accurately defined resulting in 2.9% incorrect and 64.1% false negative findings, especially synthetic inlays were hardly or not recognizable at all. For the identification of the dead especially in case of disasters with large numbers of victims, the visual dental record is still to be considered the gold standard. In the identification process itself, there is no room for error at all, although some non-concordant information may occur. Thus, PMCT should only be performed for identification in individual cases due to the relatively high error rate.

View details for DOI 10.1007/s00414-008-0274-y

View details for Web of Science ID 000259732400004

View details for PubMedID 18679703