Sex Estimation using measurements obtained from reconstructed computed tomography images of the pelvic bone of black South Africans

Sex estimation is a critical component of human identification in forensic investigations, particularly when dealing with recovered skeletal remains. In South Africa, where crime rates remain high, the largest population group, Black South Africans, is disproportionately affected. Existing pelvic-based sex estimation standards were developed from cadaveric measurements and may not be fully applicable to contemporary populations. This study, therefore, evaluated the reliability of these previously established standards by applying them to three-dimensional computed tomography (3DCT)-derived pelvic measurements from a contemporary Black South African sample. Five of the six previously published standards yielded low overall accuracies (55–78%), highlighting the need to develop new population-specific equations. The newly developed discriminant function analyses produced overall sex-classification accuracies ranging from 86.4% to 91.4%, while the logistic regression models achieved accuracies between 89.3% and 92.2%. 3DCT-derived pelvic measurements provide reliable sex estimation in the contemporaneous Black South African population group. The formulated equations demonstrate high overall accuracy and are therefore valuable tools for correct sex estimation.