This invention describes a computer method for precisely identifying and outlining individual ribs within medical images. It works by first locating a general rib area, then selecting a specific rib to analyze. For that chosen rib, the system uses a starting point derived from image data and a pre-segmentation step to accurately define its boundaries.
Why it matters: Since 2022, significant advancements in deep learning, particularly for medical image segmentation and semi-supervised learning, have made training robust models more efficient. These developments could simplify the complex seed point identification and hit-or-miss operations described, potentially improving accuracy and reducing the need for extensive manual labeling.
AI gives you a few directions you could take this. Pick one, and we check whether your version is different enough to patent, then write the filing.
Reinvent this with AI