Digital forensic osteology: morphological sexing of skeletal remains using volume-rendered cranial CT scans

Because of the increasing lack of recent bone collections, ethical issues concerning maceration procedures, and progress in radiological imaging techniques, computed tomography (CT) scans offer an alternative to traditional anthropological bone collection. The present study examined volume-rendered cranial CT (CCT) scans from 50 crania to morphologically evaluate sex characteristics. CCT scans were performed and scored by two teams (Teams A and B) of two examiners each (2x50=100 examinations) to evaluate the occurrence and/or absence of morphological traits. Altogether, 60 of 100 crania (31 Team A+29 Team B) crania were determined to be male, and 40 (19 Team A+21 Team B) were determined to be female when using the scoring system adapted from Knussmann. These results imply a sex determination accuracy rate of 96%. Only in one case was recalculation of weighting factors necessary to determine one additional correct classification. As a single parameter, arcus superciliaris evaluation permitted the most accurate sex determination (female, 84.2%; male, 85.5%). No significant difference in accuracy rates was observed between the two sexes (p<0.65, chi(2)=0.39, Fisher's exact test). Interobserver bias rates for both teams were very low (kappa=0.83). The present study shows that volume-rendered CCT images are suitable for the collection of data concerning morphologic sex determination of skulls. Thus, this method may be helpful in both actual forensic casework and the systematic reevaluation and improvement of classical anthropological methods and their adaptation to changing populations.