Fast R-CNN Approaches for Transforming Dental Caries Detection: An In-Depth Investigation
S Kanagamalliga, R. Jayashree, R Guna, Mausam Chouksey
Abstract
Dental caries, a progressive bacterial infection, is a leading cause of tooth loss, often due to inadequate dental hygiene. Addressing this pervasive issue is essential to mitigate various dental afflictions. This research's primary goal is to advance early-stage dental caries recognition using digital color imagery, enabling more effective treatment protocols. The proposed classification methodology holds potential for tele-dentistry, aligning with the growing field of tele-informatic oral healthcare. To achieve this, we implemented the Fast Region-based Convolutional Neural Networks (FRCNN) algorithm, a Convolutional Neural Network (CNN) variant, as our deep learning model. Rigorous training and testing were performed on a binary dataset featuring images with and without caries. Impressively, our results demonstrate a high accuracy rating of 99.13%. This success underscores the efficacy of the FRCNN model, showcasing its capacity to significantly improve dental caries classification. The research marks a pivotal stride in revolutionizing early-stage caries detection, promising profound impacts on oral healthcare practices and outcomes.