Real Time Bone Fracture Localization in Radiographs Using Deep Learning Techniques

This project presents an AI-powered Bone Fracture Detection system that analyzes X-ray images using advanced deep learning models and web technologies. The system integrates multiple object detection models, including YOLOv8, Faster R-CNN, and SSD, to accurately identify fracture regions in medical images. A Flask-based web application is developed to provide an interactive interface for uploading X-rays, selecting models, and visualizing detection results. Image preprocessing techniques such as noise reduction and contrast enhancement are applied to improve detection performance. The system also incorporates AI-generated medical reports using Gemini, offering detailed fracture analysis, severity assessment, and clinical recommendations. Additionally, features like grid overlay and confidence threshold adjustment enhance usability and precision. This solution aims to assist medical professionals by providing quick, reliable, and automated fracture detection, thereby improving diagnostic efficiency and reducing human error in radiographic analysis.