Advancing caries detection in dentistry: A narrative review of artificial intelligence applications

Charlie, Shabnam Mohammed and Panjnoush, Mehrdad and Fakhar, Hourieh Bashizadeh and pour, Daryoush Goodarzi (2025) Advancing caries detection in dentistry: A narrative review of artificial intelligence applications. World Journal of Biology Pharmacy and Health Sciences, 23 (1). pp. 258-266. ISSN 2582-5542

Artificial intelligence is rapidly transforming dental diagnostics, particularly in the detection of caries, offering unprecedented precision and efficiency compared to conventional methods. This review explores the evolution of AI applications in dentistry, highlighting how machine learning, deep learning, and computer vision are reshaping diagnostic processes. Traditional methods such as visual-tactile examinations and radiographic imaging, while fundamental, often suffer from limitations including human error, inconsistency, and difficulty in early detection. AI technologies address these challenges by offering consistent, fast, and highly accurate detection capabilities. Convolutional neural networks (CNNs) have demonstrated remarkable success in analyzing bitewing, periapical, and panoramic images, often outperforming human examiners in detecting early-stage carious lesions. Beyond radiographic analysis, AI-driven image segmentation enhances diagnostic precision by objectively highlighting affected regions, supporting clinicians in devising more tailored treatment strategies. Clinical applications already show that AI not only boosts diagnostic confidence but also improves patient engagement by providing visual explanations. Despite its promising potential, the field faces hurdles such as the need for large, diverse, and high-quality datasets, concerns about data privacy, and the necessity for rigorous validation across different populations. Ethical and legal considerations, particularly around accountability and explainability, further emphasize the need for clear regulatory frameworks. Emerging trends focus on explainable AI, multidisciplinary collaborations, and personalized AI solutions that integrate with electronic dental records, paving the way for more patient-specific care. Studies to date show that AI models can achieve caries detection accuracies exceeding 80%, with some nearing 99%, demonstrating the immense future promise of this technology. However, unlocking AI’s full potential in dentistry will require ongoing research, validation in real-world settings, and a concerted effort between dental professionals, AI developers, and regulators to ensure that AI systems are safe, reliable, and ethically implemented to enhance patient outcomes and revolutionize dental care.