Information from the abstract
OBJECTIVES: To systematically evaluate the diagnostic performance of artificial intelligence (AI) systems for detecting proximal caries lesions on bitewing radiographs. MATERIALS AND METHODS: Electronic searches were conducted in PubMed/MEDLINE, Scopus, and Web of Science for studies published between January 2000 and March 2026. Studies evaluating AI-based systems for proximal caries detection on bitewing radiographs were included. True-positive, false-positive, true-negative, and false-negative values were extracted directly or reconstructed from reported diagnostic accuracy metrics when sufficient information was available. Pooled sensitivity and specificity were estimated separately using random-effects models on the logit scale. RESULTS: Eleven studies were included in the qualitative synthesis, of which three studies provided sufficient 2×2 contingency table data for quantitative synthesis. The pooled sensitivity and specificity of AI systems for proximal caries detection were 0.73 (95% CI 0.70–0.75) and 0.92 (95% CI 0.69–0.99), respectively. Most included studies demonstrated promising diagnostic performance using deep learning architectures, particularly convolutional neural networks and YOLO-based models. CONCLUSION: AI systems demonstrated promising diagnostic performance for detecting proximal caries lesions on bitewing radiographs. However, methodological heterogeneity and limited external validation remain important limitations. Further prospective multicenter studies are needed to confirm the clinical applicability of AI-assisted caries detection systems.
Why this record is monitored
This record has an Impact Signal of 70/100 based on recency, source, collaboration, and bibliographic signals. It prioritizes monitoring and is not a judgment of research quality.
Related topics: Dental Health and Care Utilization · Dental Radiography and Imaging · Oral microbiology and periodontitis research
Thai researcher and institutional participation
Arnun Jukisalapong · Bangkok Hospital
Data limitations
This page is a bibliographic record based on abstract-level information, not a full analysis or quality assessment. Verify the DOI and original article before citation.