Desenvolvimento e validação de novas fórmulas para estimativa do sexo em adultos brasileiros: uma abordagem tomográfica envolvendo canal incisivo mandibular, tubérculos geni e segunda vértebra cervical

Abstract

The establishment of sex is extremely important and becomes one of the first steps in forensic anthropology in the search for human identification, which can be estimated by the morphological and metric characteristics of the human skull through cone beam computed tomography (CBCT). In cases of major disasters, finding a complete human skeleton becomes a challenge. The mandible is easily displaced from the skull through the temporomandibular joint and the cervical vertebrae are the best preserved in the human body, due to their favorable location, protected by the skull. The second cervical vertebra (C2) is easily identified even outside the human skeleton, due to the presence of the odontoid process, which clearly differentiates it from the other vertebrae in the spine. Therefore, it is necessary to study their dimorphic sexual characteristics separately. Furthermore, the accuracy in discriminating between sexes based on structures located in the anterior region of the mandible, such as the mandibular incisive canal (CIM) and the geni tubercle (TG), has not yet been studied, which reinforces the importance of studies in this area. The present investigation was developed in two phases: (1) development of three formulas based on linear measurements of CIM, TG and C2 in 150 men and 150 women and (2) validation of the formulas created with a different sample of 50 men and 50 women. The study used a linear regression model, Receiver Operating Characteristic curves were constructed to obtain area under the curve (AUC), sensitivity, specificity and accuracy, and discriminant function analysis was also performed. CIM showed a higher proportion of significantly higher measurements in males (82.4%) compared to TG (42.9%). The formulas for estimating sex presented, respectively, sensitivity and specificity of up to 71.3% and 80% for the MIC, and 67.3% and 72% for the TG. The CIM presented a formula for estimating sex with a greater area under the ROC curve (AUC) (0.829±0.024) than the TG (0.760±0.027). Discriminant analysis was statistically significant (p<0.001) and found a higher canonical correlation value for CIM (0.638) compared to TG (0.580). CIM showed better ability to discriminate sexes than TG and C2 showed better ability to estimate sex than CIM and TG. All C2 variables presented higher mean values in males (p<0.001). The variables maximum distance of the superior facets (DMFS), maximum width of the left superior facet (LMFS), maximum length of the right superior facet (CMFS), and maximum length of the vertebral foramen (CMFV) were those with significant collinearity coefficients. Formula based on the sum of these variables resulted in AUC=0.883±0.020 (p<0.001), sensitivity and specificity of 81.3%. Discriminant function analysis was statistically significant (p<0.001; canonical correlation coefficient 0.655), with the best parameter being the CMFV. The validated formula (DMFS×0.136 + LMFSE×0.165 + CMFSD×0.183 + CMFV×0.134) with a cutoff point of 14.35, this value is used as a parameter to estimate male (value above) and female (below) sex. Thus, it obtained AUC=0.893±0.034, sensitivity 90%, specificity 70% and accuracy 80%. Therefore, it can be concluded that the CIM, the TG and the C2 vertebra were shown to be dimorphic structures based on linear CBCT measurements of Brazilian adults.