Diagnostic applications of artificial intelligence in dental care for medically compromised patients: A scoping review

Objective: Accurate diagnosis is crucial for effective dental care in medically compromised patients. Artificial intelligence (AI) holds promise for enhancing early detection and diagnosis of oral diseases. However, its application in this vulnerable population remains underexplored. This scoping review investigates how AI is applied in dental care for individuals with chronic conditions that affect daily functioning and complicate dental management. The focus is on AI's role in early detection and diagnostic evaluation within the dental practices for medically compromised patients. Methods: PubMed, Scopus, Web of Science, and IEEE databases were searched using Medical Subject Headings (MeSH) and specific AI- and dental-related keywords. Articles were selected, charted, and analyzed using the Joanna Briggs Institute's (JBI) scoping review framework. Results: A total of 15 studies met the inclusion criteria. AI methods, such as Convolutional Neural Networks (CNNs), Artificial Neural Networks (ANNs), Random Forest (RF), and XGBoost, were utilised for diagnosing oral conditions, including periodontitis, osteoporosis, and oral cancer. While most studies reported encouraging diagnostic performance (sensitivity above 60 %, precision up to 99 %), several limitations were common, including a lack of external validation, small sample sizes, and heterogeneous reporting standards. Conclusion: AI shows significant potential in enhancing the diagnostic accuracy of oral diseases among medically compromised patients. Nevertheless, current evidence is limited by methodological weaknesses. Future research should focus on robust external validation, multicentre data collection, and standardised reporting of diagnostic performance to support clinical implementation.