Skip to Main Content
In this paper we present a fast and efficient method for generating a patient-specific femur mesh by morphing a template femur mesh of ten-node tetrahedral elements on a geometry represented in STL format (STL: Surface Tessellation Language). The morphing is constrained through a set of user defined anatomical landmark points. Our method splits the input geometries into open surfaces, and maps each surface individually on a planar parameterization. We morph the parameterizations individually using a regression model based on Radial Basis Functions (RBFs). In experiments our method shows precise results in generating new patient-specific meshes from existing models and in re-meshing.