A study on boundary segmentation and assembly of fractured bone fragments using measured point clouds

In virtual surgery, bone fragments can be repeatedly adjusted to get each fragment perfectly aligned. However, in real-life surgeries, only a few attempts can be made, and visualization may at times be difficult to achieve due to inadequate exposure. To address this issue, the paper proposes an automated method for assembling fractured bone fragments using measured point cloud. The method first segments the bone into two parts: the intact bone surface and the fractured surface. Next, it identifies edge points along the bone surface and divides them into smaller edge sequences. These sequences are then matched to reconstruct the original shape of the bone. This approach enhances preoperative planning efficiency and reduces the need for trial-and-error during actual surgery.