Sex Determination Using Craniometric Parameters: A Computed Tomography- Based Assessment

Objective: This study aimed to evaluate the effectiveness of cranial measurements obtained from computed tomography (CT) scans in determining sex. Seven linear cranial parameters were assessed to investigate their individual and combined discriminative power in sex classification within an adult Turkish population. Methods: Adult individuals of known sex were retrospectively selected from cranial CT scans archived at a tertiary care hospital. Seven craniometric parameters—including maximum cranial length, naso-occipital length, cranial base length, basion–bregma length, bizygomatic width, biorbital width, and interorbital width—were measured using multiplanar reconstructed CT images. Sex-based comparisons were performed, and a quadratic discriminant analysis (QDA) was conducted to assess classification accuracy. Results: A total of 200 individuals (100 males and 100 females), aged between 20 and 75 years, were included in the study. All cranial measurements were significantly higher in males compared to females (p < 0.001 for all parameters except interorbital width, p = 0.047). The most prominent sex-based differences were observed in bizygomatic width (132.82 ± 5.00 mm in males vs. 124.06 ± 4.90 mm in females) and maximum cranial length (182.17 ± 7.42 mm vs. 171.51 ± 6.86 mm). The QDA model achieved an overall classification accuracy of 86%, with 83% accuracy in males and 89% in females. Interorbital width was the only parameter showing significant variation by age group (p = 0.002), while the remaining measurements were unaffected by age. Conclusion: Cranial parameters obtained from CT imaging demonstrated a classification accuracy of 86% in sex estimation within our sample, supporting their utility in forensic anthropological assessments. These findings reinforce the value of CT-based linear craniometric analysis as a non-invasive and reproducible approach, particularly in settings where advanced 3D or AI-based methods may not be feasible. Additionally, this study provides reference data specific to an adult Turkish cohort, contributing to population-specific standards in the literature.