Morphing algorithm for building individualized 3D skeleton model from CT data

There are constant efforts to improve speed and automation of the segmentation process of CT data. We propose a semi-automatic, atlas-based method of building 3D individualized skeletons. The intervention of human operator is minimal and is limited to selection of one voxel for each bone to be segmented, measuring a few distances for a successful reconstruction and approving the final result of an alignment. As a proof of the concept, an example of creating a patient-specific model of a left femur bone is presented. The results are discussed in order to chart the route to an automatic solution. It is believed that the areas like orthopedic operation planning and patient positioning systems could benefit from our method.