The Accuracy and confidence in human dental identification using panoramic radiographs: the role of observer experience

impact of examiner experience on accuracy and confidence

Abstract

Swift and accurate disaster victim identification (DVI) is essential for medicolegal closure and repatriation of remains. Dental comparison is one of three primary identification methods recognized by The International Criminal Police Organization (INTERPOL), and is conducted by forensic odontologists, often by matching antemortem and postmortem intraoral radiographs. However, obtaining postmortem intraoral images can sometimes be challenging (e.g., from burn victims) and in those cases extraoral methods, especially computed tomography (CT), could be an alternative. This study uses conventional panoramic radiographs (PRs) to simulate CT-reconstructed PRs which are increasingly used in forensic scenarios, where traditional intraoral postmortem imaging is challenging. The aim is to assess the accuracy in comparative dental identification by different professionals when comparing antemortem intraoral radiographs to postmortem PRs.

In this retrospective study, intraoral radiographs and PRs from 30 patients were used to simulate a closed disaster event with 25 deceased and 30 possible missing persons. Twenty-five observers from four professional groups—eight oral radiology specialists (ORs), three forensic odontologists (FOs), six dentists trained in the basics of DVI (DVI-D), and eight dental students (DS) - matched PRs, serving as simulated postmortem radiographs, to antemortem intraoral radiographs using a comparative method. Statistical analysis was performed using chi-square and Kruskal-Wallis tests.

FOs and ORs achieved 100% accuracy, while DS and DVI-D had accuracy rates of 98.5% and 94.7%, respectively. Fillings were the most commonly used radiographic feature for matching. Confidence levels, ranging from “no match” to “established”, differed, with 96% of ORs and 92% of FOs selecting established matches compared with 67% of DS and 51% of DVI-D.

Conventional PRs can be used for accurate matching in dental identification, particularly when interpreted by experienced observers. FOs and ORs had significantly higher matching accuracy and confidence, emphasizing the critical importance of observer experience; these findings support that ORs may be an asset in DVI operations.

Keywords: Panoramic Radiographs, Identification,  Forensic Odontology,  Adult,  Experience