Sex identification based on paranasal sinuses measurements using multislice computed tomography in a sample of adult Egyptians

Abstract

Abstract Background Identification is a major challenge in forensic science. Paranasal sinuses show individual and sex-related variations. The complexity of the internal architecture of the ethmoid sinus, its deep anatomical location, and relative protection from trauma and surgical alteration suggest potential forensic value, particularly in cases involving partial skeletal remains. This study evaluated sex identification using frontal, maxillary, and ethmoid sinus measurements obtained from multislice CT scans of 221 Egyptian adults (103 males, 118 females). Linear dimensions and volumes of the sinuses, along with selected inter-sinus distances and ethmoid-related measurements, were analyzed. Results The studied measurements tended to be greater in men compared to women. Multiple logistic regression models were developed for sex prediction using individual sinuses. The frontal sinus model achieved a sensitivity of 56.3% and a specificity of 72.0%, while the maxillary sinus model showed a sensitivity of 50.5% and a specificity of 76.3%. The ethmoid sinus model demonstrated higher performance, with a sensitivity of 68.9% and specificity of 83.1%. The all-sinus model yielded the best results, with a sensitivity of 72.8% and specificity of 79.7%. Overall accuracy was 64.71% for the frontal sinus, 64.25% for the maxillary sinus, and 76.47% for both the ethmoid and all-sinus models. Conclusion Paranasal sinus measurements exhibit sexual dimorphism, allowing sex prediction with moderate accuracy. Among individual sinuses, the ethmoid sinus showed the highest discriminatory ability, while combining all sinuses further improved predictive performance. Future studies should validate these models in larger, diverse populations.