Automatic Femur Decomposition, Reconstruction and Refinement using Superquadric Shapes

R. Cuypers, S. Kiel, and W. Luther (Germany)


Surgery simulation, biomedical engineering, superquadrics, geometrical refinement, femur modeling, validation


In this paper we present a fully automated method for the decomposition and reconstruction of the human femoral bone from point clouds extracted by CT (Computer To mography) or MRI (Magnet Resonance Imaging). Local and global surface analysis techniques are used to create an initial clustering of the data, which helps in selecting the appropriate geometry types for the areas in question. Components are partially fitted and reintroduced to the reconstruction process to retrieve other components, which in turn help improve the quality of the previous fits. The model is refined to optimize the number and spatial ordering of the model parts. Finally, a connection is estab lished to a semi-automated approach along with a qualita tive comparison. Another issue is the validation of meas urements performed on superquadric-based models, which is discussed in the context of THA (Total Hip Ar throplasty) surgery.

Important Links:

Go Back