Artificial Intelligence & Deep Learning in Forensic Odontology: From Automated Age-Estimation to Dental Identification

Abstract

Introduction: Forensic odontology plays a pivotal role in human identification, age estimation, and bite mark analysis. The rapid integration of Artificial Intelligence (AI) and Deep Learning (DL) has revolutionized data-driven forensic workflows, offering enhanced accuracy, speed, and reproducibility.
Material and Method: To systematically map and synthesize the scope, applications, and trends in AI and DL for forensic odontology, from automated age estimation to dental identification the scoping review was done following the PRISMA-ScR guidelines. Searches were conducted across PubMed, Scopus, Web of Science and Google Scholar up to August 2025 using keywords “Artificial Intelligence”, “Deep Learning”, “Forensic Odontology”, “Age Estimation” and “Dental Identification”. Eligible studies included original research, reviews and reports in English language involving AI/DL applications. Data were charted and synthesized accordingly.
Results: From 486 initial records, 26 met inclusion criteria. Convolutional Neural Networks (CNNs) were the most utilized architecture for dental radiograph analysis. Applications were clustered into: (i) Automated age estimation , (ii) Sex determination, (iii) Dental identification and (iv) Bite mark pattern recognition. Reported accuracies for age estimation ranged from 85% to 97%.
Conclusion: Transfer learning and hybrid models showed increasing adoption post-2020. AI and DL show high potential in forensic odontology, with CNN-based models dominating the landscape. However, challenges in dataset standardization, model validation across populations, and legal admissibility remain. Future research should focus on explainable AI and cross-population generalizability.