Sensitivity and specificity of DentMA teledentistry in dental caries detection: screening in vulnerable communities

Abstract

Introduction: Teledentistry offers a practical solution to limited access to dental care by enabling remote screening using smartphone-based technology. Vulnerable populations such as pregnant women, children who rely on parental assistance, and the elderly have a high prevalence of dental caries and often face barriers to routine dental visits due to mobility limitations, time constraints, and dependence on caregivers. These characteristics make them priority targets for teledentistry-based screening. DentMA teledentistry is a smartphone-based application developed by the research team, integrating intraoral photographic capture, deep learning–assisted caries detection, and remote dentist evaluation to support early dental screening. This study aimed to evaluate the sensitivity and specificity of DentMA teledentistry for dental caries screening in vulnerable community populations.  Methods: This was a prospective study with a cross-sectional design. Intraoral photographs were obtained using a smartphone camera and then uploaded to an Internet of Dental Things (IoDT)-based platform. Intraoral images were reviewed by dental professionals and compared with clinical examinations by co-assistants and dental nurses during activities outside the health center. Photographic assessments carried out by dentists were compared to clinical assessments by co-assistents and dental nurses. Data analysis included sensitivity, specificity, and accuracy, accompanied by Cohen’s kappa values >0.60 Results: 90 participants were included in this study, consisting of 30 pregnant women, 30 elderly individuals, and 30 children. All photographic views showed high sensitivity of 85.54% (95% CI, 77.0% to 93.1%) and specificity of 86.60% (95% CI, 79.8% to 93.4%) in detecting email-dentin caries and PUFA, with an accuracy of 86.11%. The agreement between the results of clinical examinations and photographic methods (assessed by dentists) was considered strong, with kappa scores ranging from 0.74-0.80. Conclusion: Assessment of intraoral photographs using DentMA teledentistry for caries detection demonstrated clinically acceptable accuracy when compared with visual clinical examination, based on sensitivity and specificity values. This strategy has implications for supporting remote dental health services for vulnerable communities (pregnant women, the elderly, and children) without access to healthcare facilities.