Researchers have developed an innovative artificial intelligence system that can detect, classify, and segment hip fractures in x-ray images with high precision. Trained on thousands of annotated radiographs, the model identifies the exact location and type of fracture—whether in the femoral neck, pertrochanteric, or subtrochanteric region—and grades its severity using established medical scales. Unlike previous methods that only labeled entire images, this approach highlights the specific fracture area, offering clearer visual cues for diagnosis. With an AUC of over 96% in fracture detection and 86% in classification, the system shows promise as a valuable tool for improving diagnostic confidence and supporting clinical decision-making.
