Computed tomography as a solution to skeletal collection limitations : exploring cranial variation in the South African population

Abstract

Skeletal collections have provided the foundation for forensic anthropology in South Africa, offering invaluable reference material for developing and testing identification methods. However, these collections remain limited in both size and representativeness. Many are historically biased, restricted by donation practices, and unable to reflect the demographic and biological diversity of the country’s modern population. Indian South Africans, in particular, are notably absent due to cultural and religious beliefs that discourage body donation. This study explores the use of computed tomography (CT) as an alternative to overcome these limitations. A total of 408 high-resolution head CT scans representing Black, Coloured, Indian, and White South Africans were analysed using inter-landmark distances and geometric morphometrics. Measurement repeatability was high, with both intra- and inter-observer errors falling well within accepted forensic thresholds (< 2 mm; <3.5%). Significant population-level differences in cranial form were identified (p < 0.01). Black South Africans were consistently the most clearly differentiated group, while Coloured and Indian South Africans showed the greatest overlap, reflecting shared population affinity and historical admixture. Cross-validated classification accuracies ranged between 62% and 64%, indicating that both linear and shape-based analyses can capture relevant cranial variation. These results confirm that CT-derived data are reliable for examining cranial variation within contemporary South African samples and provide access to larger and more diverse datasets than traditional skeletal collections. This study establishes the first cranial reference data incorporating Indian South Africans and demonstrates how CT scanning can support the development of inclusive, population-specific standards for forensic anthropology in South Africa.