Evaluation of a Smartphone-Based Caries Detection App in Rural Dental Camps

Aim: To evaluate the diagnostic accuracy and feasibility of a smartphone-based caries detection application in rural dental camps, comparing its performance against clinical and radiographic standards. Methodology: A prospective diagnostic-accuracy study was conducted across 12 rural camps, enrolling 582 participants (≥5 years). Standardized intraoral images were analyzed by the app’s AI model and compared with ICDAS-based examinations by calibrated dentists. Results: The app's overall accuracy was 85.4%, its per-tooth sensitivity was 87.2%, its specificity was 83.1%, and it had a significant agreement (κ=0.72) with clinicians. Occlusal caries performed better (90.3%) than proximal (78.9%), and the feasibility results showed few problems with connectivity and image rejection. Conclusion: The smartphone app's potential for widespread caries screening in underserved rural areas was supported by its high diagnostic accuracy, quick processing, and strong agreement with clinical gold standards.