EMRE SÖYLEMEZ SUNA TOKGÖZ YILMAZ
Northwestern Medical Journal - 2025;5(2):77-84
Aim: The pinna, the hearing organ, also contributes to the aesthetic appearance of the face. We aimed to investigate the feasibility of sex prediction using anthropometric measurements of the pinna in machine learning. METHODS: The study included two hundred healthy individuals (104 women and 96 men). The pinna of these individuals were measured in eight different parts using a digital calliper. The data, which differed by sex, were processed in eight different machine-learning algorithms. Results: Seven different measurements, such as pinna length, width and lobule length, were greater in men than in women (p<0.05). The K-Nearest Neighbor model showed the best success in sex prediction with an accuracy of 0.825 and a ROC value of 0.882. Conclusions: Pinna’s anthropometric measurement values can be used in machine learning to predict sex with a high success rate. Our study shows that ear prints may have potential use in forensic identification.
