Elastic deformation for automated planning of surgical interventions

Abstract In the frame of the collaborative research centre “Computer and Sensor Aided Surgery” (SFB 414), we developed a planning environment called “KasOp”, which supports the surgeon to define osteotomies in cranio–maxillo-facial surgery. However, planning of surgical interventions still requires a lot of user interaction in order to precisely define cutting trajectories. The presented approach in this paper uses rigid and elastic deformation of polygonal surfaces in order to align pathologically shaped patient bone surfaces and predefined reference bone surfaces of a physiological data set of same age and sex. By doing this, we are able to transfer a trajectory predefined on the physiological data set onto the patient without any user interaction. This speeds up the planning process, as the surgeon simply needs to modify a preplanned trajectory instead of creating a new one. Moreover, the surgeon receives an impression of how the patient should look like, and regions with the largest deviations to the physiological shape can be identified.