Gender Prediction Using Machine Learning Algorithms with Parameters Obtained from Calcaneus

Background: Partial or total disruption of body integrity may occur in cases such as war, natural disasters and accidents. In such cases, the importance of the calcaneus bone, which has a hard and minimal structure, increases for identification. With this hypothesis, the aim of this study is to estimate gender from the calcaneus by utilising the current approach of machine learning (ML) algorithm. Materials and Methods: The study was performed on X-Ray images of 200 female and 200 male subjects aged 18-65 years. Maximum length, facies articularis cuboidea height, maximum width, body width, minimum length, anteroposterior length of the calcaneus, posterior facet angle, anterior angle of the cuboid facet of the calcaneus, facet height, posterior facet length, anterior proces length, calcaneus inclination angle, talocalcaneal angle, Böhler angle, Gissane angle and calcaneus tuber angle were measured. Then the obtained data were used in the input of ML algorithms. Results: As a result, a highly accurate and reliable sex prediction rate between 0.86-0.91 was obtained with ML algorithms. In addition, it was found that the maximum width of the calcaneus parameter made the highest contribution to sex prediction among the parameters with SHapley Additive exPlanations. Conclusions: As a result of our study, it was found that calcaneus with minimal and rigid structure provided high accuracy in terms of gender prediction using ML algorithms. In this respect, we think that this study will be a reference for forensic and morphometric studies.