Age Estimation Based On Orthopantomograms Using Artificial Intelligence

This study evaluated the accuracy of artificial intelligence (AI) in determining age based on orthopantomograms (OPG), considering gender differences. The research utilized the DASE AI tool (Dental age and sex estimation by artificial intelligence) developed by the University of Zagreb Faculty of Electrical Engineering and Computing and the School of Dental Medicine. The study examined the OPGs of 700 individuals (398 females, 302 males) aged 20-71 years (Croatian Caucasian population). Inclusion criteria were known date of birth, date of OPG taken, sex, and at least 16 visible teeth. AI-predicted ages were compared to actual ages (Pearson's correlation coefficient). Mean absolute error (MAE) was calculated to evaluate prediction accuracy. Gender-based differences were analyzed using independent t-tests. AI-predicted ages moderately correlated with actual ages (r = 0.65, p < 0.001). The overall mean absolute error (MAE) was 4.7 years. AI predictions were more accurate for individuals under 50 (MAE: 4.0 years) than those over 50 (MAE: 7.1 years). Gender-based analysis revealed slightly higher accuracy for males (MAE: 4.3 years) than females (MAE: 5.0 years), though this difference was not statistically significant (p > 0.05). AI demonstrates potential in age determination from orthopantomograms, with moderate accuracy that varies across age groups. While slight gender-based differences in prediction accuracy were observed, they were not statistically significant. Further refinement of AI algorithms and analysis of larger datasets may improve prediction accuracy, particularly for older individuals.