Detection, Segmentation and Classification of Tooth Cavities based on Medical Records

This paper presents a comprehensive study on the detection and segmentation of tooth cavities and the classification of dental anomalies using medical records. The primary objective was to develop an automated, efficient, and accurate system for dental diagnosis, addressing the limitations of traditional diagnostic methods. The core of the work entailed designing and implementing classical MLA, deep learning algorithms, particularly YOLOv8 and U-net, for the analysis of dental images. These algorithms were chosen for their proven effectiveness in object detection and image segmentation in medical imaging. The study involved training these models with a substantial dataset of annotated dental X-ray images and corresponding medical records. The results showed a significant improvement in the detection and segmentation of tooth cavities over traditional methods. YOLOv8 and U-net demonstrated high accuracy and efficiency in identifying dental anomalies. Furthermore, the integration of patient medical records contributed to a more nuanced understanding and classification of dental conditions. The findings of this study underscore the potential of artificial intelligence and deep learning in revolutionizing dental diagnostics. This research contributes valuable insights into the application of AI in dental healthcare and lays a foundation for future advancements in the field.