Evaluation of the Use of Artificial Intelligence in Dental Imaging Interpretation: A Systematic Review

The rapid adoption of digital radiography in dentistry has generated large volumes of imaging data, creating opportunities for artificial intelligence (AI)-based tools to assist in diagnostic interpretation. AI, particularly machine learning and deep learning models, has shown promising applications in improving diagnostic efficiency, consistency, and image analysis in dental imaging. This systematic review aimed to evaluate the applications, diagnostic performance, and limitations of AI in dental imaging interpretation. A systematic literature search was conducted in accordance with Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines using PubMed (MEDLINE) and ScienceDirect databases for studies published between May 2005 and March 2024, employing relevant MeSH terms related to artificial intelligence and dental imaging. Following database-specific search refinement and application of predefined eligibility criteria, 25 studies were screened, of which seven met the final eligibility criteria and were included in the qualitative synthesis. The findings indicate that AI applications in dental imaging, including caries detection, periodontal bone loss assessment, impacted third molar evaluation, root fracture detection, working length determination, and image registration, demonstrated supportive diagnostic performance and improved efficiency compared with conventional interpretation methods. Although AI systems are not a replacement for clinical expertise, they may serve as useful decision-support tools in dental radiology. Overall, within the limited evidence base included in this review, AI-assisted tools showed potential to support diagnostic workflows and image interpretation in dental radiology. However, because only a small number of heterogeneous studies were included, these findings should be interpreted cautiously, and further high-quality research and external validation studies are required before widespread clinical implementation.