Determining the Age Range Based on Machine-Learning Methods From Facial Skeletal Angles (Glabella and Maxilla Angle and Length and Width of Piriformis) in CT Scan

Background: One of the main steps in identifying a person in forensic medicine is determining the age of skeletal remains, including the skull. This study aimed to investigate the possibility of predicting age from facial angles (glabella, piriformis, and maxillary angle and measuring peripheral length and width) with artificial intelligence in a CT scan. Methods: The cross-sectional study method is simple random sampling using a questionnaire. Accurately measurable CT scan samples are selected. For exclusion criteria, gender uncertainty, and the possibility of measurement based on CT scan quality, the researchers examined the facial angles (angle of the glabella and maxilla and length and width of the piriformis) for 100 men and 100 women. The Mean±SD of the age was 39.16±2.22 years for men and 47.84±2.46 years for women. The samples were classified based on age differences, and then the data were analyzed using machine learning algorithms to determine the age group. Results: After determining the exact amount of measurement, the data were evaluated by machine learning algorithms to determine the age group. Accordingly, in the age group classification based on the World Health Organization (WHO) (with an age difference of 10 years) (years±5) with 100% accuracy and in the second classification (with an age difference of 5 years) (years±2.5) with 88% accuracy and 79% precision of the age group was predicted. Conclusion: The obtained data show the importance of new artificial intelligence methods, including machine learning, in providing new methods to determine age groups (age±2.5) through skull angles with high accuracy in cases where even cranial remains are found in identification in forensic medicine.