Análise da predição sexual utilizando ossos nasais e abertura piriforme por meio de tomografia computadorizada de feixe cônico : um estudo retrospectivo com a população da Região Sul do Brasil

Abstract

Introduction: Human identification is crucial in cases where only skeletal remains are recovered, such as accidents, explosions, fires, and natural disasters. Forensic dentistry contributes to this process through craniometric analyses and morphological characteristics to estimate chronological age, sex, ancestry and stature. Aim: To analyze the accuracy of nasal bone and piriform aperture measurements using Cone-Beam Computed Tomography (CBCT) images for sex prediction in individuals from the southern region of Brazil, through the application of artificial intelligence. Methods: A sample of 271 CBCT images (168 female and 103 male) was selected from a dental radiology clinic in southern Brazil, comprising patients aged 20 to 60 years. Measurements of the two anatomical structures were performed by a single examiner from different views. The collected data were tabulated, and a Support Vector Machine (SVM) classification algorithm was applied. Results: The SVM model achieved an accuracy of 74% when trained using the height and width of the piriform aperture and the width of the nasal bones at their most inferior portion. Regarding nasal bone shape, type A was observed in 40.59% of the population and type B in 40.96% of the analyzed sample. Conclusions: The analysis of nasal bone shape revealed a heterogeneous distribution within the southern Brazilian population. Measurements of the piriform aperture and nasal bones particularly the width at the most inferior portion show potential for contributing to sex prediction. [resumo fornecido pelo autor]