Development and validation of an age estimation model based on dental characteristics using panoramic radiographs

Dental characteristics have considerable potential as indicators for estimating chronological age. This study developed a regression model for age estimation using dental characteristics observed in panoramic radiographs. A total of 2,391 radiographs from individuals aged 20 to 89 years were analyzed, with a focus on five treatment-induced characteristics. Analyses revealed statistically significant correlations between all observed characteristics and chronological age, supporting the potential of dental characteristics as novel age indicators. A model incorporating only posterior teeth from both jaws achieved an adjusted R-squared value of 0.564 and a root mean square error (RMSE) of 13.144 years, closely comparable to the full-dentition model, which had values of 0.558 and 13.235 years, respectively, and is regarded as the most recommendable model for practical application. On the same test set, the developed model had an RMSE that was 2.651 years higher than that of a non-destructive method widely used in Korean forensic practice, indicating slightly lower accuracy. Nevertheless, given its convenience in forensic practice, it should be considered a supplementary tool used alongside conventional methods rather than a direct replacement. Future research should leverage diverse datasets from multiple institutions and apply advanced technologies, such as machine learning, to enhance the applicability and robustness of dental characteristic-based models.