Leveraging 3D Faster R-CNN for 3D Dental X-ray Restoration and Treatment Identification

This research aimed to develop a 3D Faster-R-CNN model for detecting dental restorations and treatments in panoramic view radiographs and dental intra-oral X-rays. Trained on a comprehensive collection of 2D and 3D dental X-ray images, it marked a notable advancement. Utilizing the 3D TeethSeg 2022 MICCA Dataset, the model demonstrated its ability to automatically detect and classify dental features, including challenging traits like cavity spots and tooth cracks. The study addressed the limitations of current methods for analyzing dental X-rays, especially in 3D format. While existing 2D models have some recognition capabilities, they struggle with 3D data intricacies. Moreover, dedicated 3D models for dental applications are lacking. This proposed model introduced a novel approach tailored for 3D dental image analysis, aiming to fill this gap. The model played a vital role in assisting dentists by identifying restorative and treatment needs for each tooth or patient, significantly enhancing dental care quality and efficiency. It showcased notable performance metrics: an accuracy of 89.14%, a precision of 95.84%, and a recall of 91.93%. Through advanced data augmentation and meticulous model training, it represented a significant advancement in detecting and classifying dental interventions, offering a comprehensive approach to dental image analysis.