Alveolar Bone Loss Detection and Localization in Dental X-Ray Images using YOLOv5

Abstract

Periodontal disease, characterized by alveolar bone loss, is a prevalent oral health condition that requires early detection and management to prevent further progression. This paper proposes a novel approach for alveolar bone loss detection and localization in dental X-ray images using the YOLOv5 object detection algorithm. We annotated a dataset of dental radiographs with alveolar bone loss regions and fine-tuned the YOLOv5 model on this dataset. Our approach achieved high accuracy and robustness in detecting and localizing alveolar bone loss regions, with precision, recall, and F1 score exceeding 90%. The real-time processing capabilities of YOLOv5 make it suitable for clinical implementation, providing an efficient and accurate solution for periodontal disease management. The automated alveolar bone loss detection and localization using YOLOv5 can significantly assist dentists in the early diagnosis and treatment planning of periodontal diseases, leading to improved patient outcomes and reduced risks of tooth loss. The proposed method has the potential to be integrated into clinical practice, providing a valuable tool for dental practitioners in the field of periodontics.