Objective comparison of 3D dental scans in forensic odontology identification

In forensic odontology disaster victim identification, it is crucial to assess the similarity between post mortem (PM) dentitions and ante mortem (AM) dental records from a database. To facilitate ranking AM records by likelihood of a match, the similarity evaluation must yield an intuitive, quantitative score. This study introduces a scoring scheme designed to effectively distinguish 3D dentition surfaces acquired by intraoral 3D photo scans. The scoring scheme was validated on an independent dataset. The scoring scheme presented utilizes two levels of surface similarity, spanning from local similarity of surface representations to regional similarity based on relative keypoint placement. The scoring scheme demonstrated exceptional discriminatory power on the validation data, achieving a Receiver Operating Characteristic (ROC) area-under-the-curve (AUC) of 0.990 (95% CI 0.988 to 0.992). Implementing such a scoring system in disaster victim identification workflows, where AM 3D data can be procured, can provide an initial likelihood of matching, enabling forensic professionals to prioritize cases and allocate resources more efficiently based on objective measures of dental similarity.