A Survey Analysis on Dental Caries Detection from RVG Images Using Deep Learning

A rapid developing area of study in the healthcare industry is medical imaging. Medical imaging is crucial for the early identification, diagnosis, and treatment of disorders. Ultrasound, magnetic resonance, computed tomography, X-rays, and other imaging techniques fall under this category. Dental caries is one of the most common conditions affecting people of all ages globally. Using radiovisiography (RVG) or dental X-ray images, it could be challenging to spot dental caries in its early stages. Practically, all medical fields use deep learning to predict or identify certain diseases. This research work has proposed and investigated a customized convolution neural network (CCNN) for caries detection from dental X-ray images. Few datasets are available for dental image processing. A total of 1300 X-ray images, 500 intra-oral photographic images, and 150 OPG images were collected from two medical practitioners with ground truth and used as a dataset for dental disease detection and image segmentation. This research work has demonstrated that image enhancement techniques play an important role in improving the quality of dental radiograph images. This research work explores different image augmentation techniques, which contribute to increasing the size of the dataset and improving accuracy. Image segmentation plays an important role in further classification and dental disease detection. Various deep learning models, including AlexNet, EfficientNetB0 and B7, Inception3, VGG-16, and the proposed CCNN model, are used to train the dataset. With more accuracy, a comparison between the deep learning model and the suggested CCNN model has been conducted.