Dental Fluorosis Analysis: A Web-Based Dental Fluorosis Severity Detection

The success of deep learning algorithms in biomedical applications is notable, yet their application to dental hard tissue conditions, including Dental Caries, Dental Abrasion, Dental Attrition, Erosive Tooth Wear (ETW), and Fluorosis, remains underexplored, particularly in analyzing photographic images. Traditional visual examination methods for clinical diagnostics are hindered by subjectivity. With the advancement of computer technology, there is growing interest in leveraging Artificial Intelligence (AI) for dental health. Our study in Vijayawada revealed a lack of oral health awareness among residents. Collaborating with Dr. Sandhya, we identified a significant issue—dental fluorosis affects nearly 45% of residents, with many cases going undiagnosed. Early detection and intervention could potentially reduce severity by up to 70%. This paper introduces a comprehensive solution for analyzing dental fluorosis. Utilizing Convolutional Neural Network (CNN) and You Only Look Once (YOLO) architecture, our web application swiftly detects severity. The CNN model, trained on diverse data, accurately classifies severity levels, and YOLO ensures real-time analysis. Users, including patients and dental professionals, interpret severity classifications through an interactive map interface. The final classification model achieves an impressive 96% accuracy in categorizing dental fluorosis severity and an 84% precision in object detection. This convergence of advanced image detection algorithms and user-friendly interfaces marks a significant stride in advancing dental health initiatives, fostering a healthier, more informed society in Vijayawada, and beyond.