Russian Scientists Use Neural Networks to Design Bone Implants
Researchers at Kurgan State University are developing an AI system that can isolate bone tissue from CT scans with 98 percent accuracy in under a minute, paving the way for personalized implants.
Scientists at Kurgan State University are turning to neural networks to design custom bone implants. According to the university’s press office, the AI was trained to identify bone tissue from CT scans in less than a minute.
It’s a team effort: doctors visually mark the implant’s boundaries, while the AI generates a personalized 3D model. These implants are used to repair trauma damage, congenital defects, or cancer-related bone loss — cases where replicating the geometry of a patient’s bone is critical. To train the neural network, researchers semi-manually processed about 5,000 CT images. Today, the system achieves 98 percent accuracy.
The next step is to build a platform that integrates automated image segmentation, pathology detection, and implant modeling.
