Bone Fracture Detection Using YOLOv8 and Grad-CAM Visualization

Abstract

An intelligent web-based tool was created to help medical professionals quickly and accurately identify bone fractures from X-ray images. This web app can quickly and accurately identify a variety of fracture types, including transverse, oblique, spiral, comminuted, and greenstick fractures, by incorporating cutting-edge deep learning techniques, especially the YOLOv8 object detection model. Grad-CAM visualization, which identifies the precise areas of the image that affected the model's predictions, is incorporated into the system to improve clinical trust and interpretability. Along with comprehensive analytical feedback like confidence scores and fracture type annotations, the platform provides an easy-to-use web interface for uploading images and interpreting results. Furthermore, electronic medical records and hospital information systems can be seamlessly integrated with RESTful API endpoints. Our project seeks to facilitate quicker clinical decision-making.