Sex estimation standards from CT-derived humeral measurements in Black South Africans

Anthropological data collected across antemortem, perimortem, and postmortem contexts are central to human identification, with sex estimation serving as the most critical initial step. Accurate sex determination can reduce the pool of possible identities by up to 50%. In South Africa, where crime disproportionately affects Black South Africans, prior humeral sex estimation studies have relied on calliper-based osteometric measurements of postmortem remains, often from older individuals of European descent. This study employs computed tomography (CT) to generate contemporary, population-specific data for Black South Africans and evaluates the applicability of previously published humeral sex estimation equations. Humeral measurements were obtained from CT scans of 97 individuals (33 females, 64 males) archived in the Radiology Department of Charlotte Maxeke Johannesburg Academic Hospital. Three-dimensional reconstructions were created using Xiris and IntelliSpace software, from which multiple measurements were recorded. Application of existing equations to this dataset yielded low classification accuracy, necessitating the development of new predictive models. Discriminant function analysis produced high sex classification accuracies, ranging from 83.5% to 90.7% for multivariate functions. These findings demonstrate that CT-derived humeral measurements from living individuals provide a promising alternative for sex estimation than traditional calliper based methods and highlight the importance of population-specific data in forensic anthropology.