Zhaokang Du; Pagorn Navic; Apichat Sinthubua; Patison Palee; Phannavich Malawan & Pasuk Mahakkanukrauh

As computed tomography (CT) becomes increasingly integrated into forensic casework, metric-based sex estimation from the skull remains is a fundamental approach for biological identification. Moreover, the comparability of measurements obtained from skeletal specimens and those derived from three-dimensional (3D) CT reconstructions—and the stability of sex-estimation accuracy across measurement modalities—require further evaluation. Therefore, this study aimed to develop population-specific sex estimation models in a contemporary Thai population based on 3D-CT measurements and evaluate measurement agreement between 3D-CT and dry skull measurements. All samples were obtained from the Osteology Research and Training Center (ORTC), Faculty of Medicine, Chiang Mai University. A total of 252 individuals (126 males and 126 females) were used to develop the sex-estimation equations, and an independent holdout sample (n = 26) was used to validate the accuracy. A total of 31 measurements were obtained from both 3D-CT skull images and dry skulls. Stepwise multivariate discriminant function analysis was applied to derive sex-estimation equations, and cross-modality performance was assessed by substituting dry skull measurements into CT-based equations. The results showed that sex estimation accuracy ranged from 84.62 % to 92.31 %, with combined cranial–mandibular models demonstrating the highest accuracy. When dry skull measurements were applied to CT-derived discriminant functions, classification accuracy remained comparable across measurement modalities. Intra- and inter-observer reliability were high for both modalities (R > 0.90). Furthermore, the intra-class correlation coefficient (ICC) demonstrated excellent inter-modality agreement (0.970–0.999), with minimal absolute differences between dry skull and 3D-CT measurements. Overall, the results indicate that sex estimation remains consistent across two modalities. The high level of agreement and strong reliability support the use of CT-based osteometry as a non-invasive, standardized, and reproducible method for sex estimation in forensic contexts.

KEY WORDS: Computed tomography; Skull; Mandible; Sex estimation; 3D.

DU, Z.; NAVIC, P.; SINTHUBUA, A.; PALEE, P.; MALAWAN, P. & MAHAKKANUKRAUH, P. Sex estimation from skull measurements using a three-dimensional computed tomography in a contemporary Thai population: Agreement, reliability, and comparative performance with dry skull. Int. J. Morphol., 44(4):1424-1432, 2026.