Xray Vision Solutions

We built an AI-based medical imaging tool that detects fractures on human X-rays and highlights the area of the anomaly. The system also includes a smart medical assistant that generates contextual responses based on the fracture prediction and confidence score. The goal is to support doctors and radiologists with faster and more reliable fracture diagnosis. The tool was built using a convolutional neural network classification model trained on labeled X-ray images. For localization, we integrated SHAP to generate explainable heatmaps. A Streamlit frontend connects to a FastAPI backend, and OpenAI's API powers the assistant’s medical recommendations. Next steps include expanding the dataset, and improving localization precision.