Decedent positive identification via visual comparisons of frontal sinus radiographs is commonly used in the medicolegal field; however, only a handful of studies have empirically tested this method. This study aimed to test the accuracy of visual assessment in frontal sinus identifications across a large and experientially diverse participant sample. A Qualtrics survey presented participants with 25 pairs of cropped frontal sinus radiographs, asking them to determine if they matched and their confidence level. Radiographs were from the American Association of Orthodontics Legacy Collection. Eighteen radiographic pairs were of the same individual taken a year or more apart. Seven pairs were from different individuals (nonmatches). Euclidean distances were used to select challenging nonmatches with similar outlines. Participants were also asked questions about their profession, training, and experience. The overall accuracy of the 145 respondents (3625 comparisons) was 89.9%, with a median accuracy of 92.0%. The majority of respondents (64.58%) report zero radiographic identification experience. Incorrect responses were biased, with only 3.6% of nonmatches wrongly reported as matches (false positives). Statistical analyses revealed significant effects of profession, radiographic experience, and training on match accuracies and confidence levels (p < 0.05), with a significant correlation between accuracy and confidence level (rs = 0.302, p < 0.001). These results support the use of frontal sinus visual comparisons in forensic identifications but highlight the importance of training and experience. In practice, accuracy rates are expected to exceed those reported here, given that identifications are made by medicolegal personnel using higher quality radiographs of the entire cranium.
Keywords: RADid; forensic anthropology; frontal sinus; personal identification; radiographic identification; visual assessment.
© 2025 American Academy of Forensic Sciences.