Variable selection in gender and age decision-making for traumatic spine and thoracic pathologies after various accidents with Multivariate Adaptive Regression Spline (MARS)

Abstract

Trauma is a condition that affects the body’s structure and results from outside factors. After heart disease and cancer, it is the most common cause of death across all age categories. For a variety of causes, people are routinely exposed to traumatic vertebral, thoracic pathologies and rib fractures. Ribs can be harmed by simple falls, impacts, and blunt injuries as well as broken due to car accidents and falling from a height. Magnetic resonance imaging or computed tomography are used to diagnose these fractures. In this study, non-linear complex methods were used to categorize gender and age by utilizing thoracic pathologies, fractures or cracks in the body as a result of traffic accidents or falling from a height, which have the feature of being a case in forensic issues. The most important data in the classification of gender and age were determined by Multivariate Adaptive Regression Spline (MARS) method. Although autopsy should be utilized in these situations, complex regression methods is intended to have an impact on quick and accurate decision-making about events in order to speed up or direct the process in the field of forensic medicine. As a result, the effectiveness of the experts subsequent predictions will be increased by the preliminary findings produced by real-world data and artificial intelligence algorithms or complex non-linear regression problems.