Artificial intelligence in dental age estimation- applications, technological advances and legal aspects: A narrative review.

Background: Dental age estimation constitutes a cornerstone in forensic odontology, pediatric dentistry, and medico-legal investigations. Traditional radiographic methods such as those by Demirjian, Willems, and Cameriere, though widely validated, are limited by examiner subjectivity, population-specific calibration, and low scalability. This narrative review examines the current landscape of artificial intelligence (AI)-driven dental age estimation, with a focus on deep learning technologies, comparative advantages over conventional methodologies, and applicability across clinical, forensic, and legal domains. Methods: A literature search was conducted to identify original studies and systematic reviews that employed machine learning (ML) and convolutional neural networks (CNNs) for dental age estimation using panoramic radiographs or cone-beam computed tomography (CBCT). Emphasis was placed on studies reporting model architecture, mean absolute error (MAE), classification accuracy, and external validation. Results: AI-based models, particularly CNNs, demonstrated superior diagnostic performance with MAEs ranging from 0.03 to 0.7 years and classification accuracies exceeding 90 % at critical legal thresholds. These systems provide automated tooth detection, segmentation, and staging, with outputs that are rapid, objective, and reproducible. Nonetheless, critical limitations persist, including algorithmic opacity, demographic bias due to non-representative training datasets, and absence of international validation standards. Conclusion: AI technologies represent a paradigm shift in dental age estimation, offering enhanced precision and operational efficiency. To facilitate clinical translation and forensic admissibility, future efforts must prioritize population-diverse training datasets, transparent algorithmic design, and consensus-driven regulatory frameworks.