AI-driven 3D wisdom tooth models for age estimation: Novel pictorial tables for caucasians aged 16–22 years

Objectives: Dental age estimation has a crucial role in forensic, clinical and anatomical contexts. This study introduced a novel approach using AI-generated three-dimensional (3D) pictorial tables based on third molar development, transforming traditional two-dimensional (2D) methods, Demirjian's and Köhler's, into an 3D format. Methods: This study included 140 Caucasians aged 16–22 years (10 males/females per age). Each subject had a panoramic radiograph and CBCT scan acquired within 30-day. Two independent examiners assessed the panoramic radiographs using the aforementioned 2D age estimation methods, a third examiner consulted in disagreement. Representative 3D models based on CBCT were generated and used to construct pictorial tables. Although 3D models were not scored directly, they functioned as anatomical references to enhance interpretation of 2D assessments and establish a framework for future integration of 3D tools in age estimation. Results: Age estimation based on panoramic imaging revealed recognisable developmental patterns using both methods. However, distinguishing between younger and older than 18 years remains challenging. Despite this limitation, certain stages were predominantly observed in individuals over 18 years. AI-generated 3D models improved wisdom tooth visualization, thereby eliminating overlap and highlighting root development for enhanced tooth maturity and age estimation. The 3D pictorial tables depicting age-related AI-wisdom tooth models showed promising potential for age estimation. Conclusions: This study demonstrates the feasibility of using AI-driven 3D models to improve age estimation in the Caucasian population through pictorial tables based on third molar development. Although promising, further validation is required to confirm the accuracy compared to traditional panoramic imaging.