ABSTRACT
Bone is nonhomogeneous, and its material properties vary across the whole bone. In this chapter, two approaches were applied to study nonhomogeneous bone. In the first method, the finite element model was built based on computed tomography (CT) data of bone. This method has two drawbacks, however. One is that CT data has noises, which makes the material properties of bone inaccurate. The other is that purely empirical equations are used to convert CT data to material properties of bone. The second approach attempts to avoid these issues by interpolating the material properties of bone from the experimental data using multidimensional interpolation algorithms, such as the Radial Basis (RBAS) algorithm, Nearest Neighbor (NNEI) algorithm, and the Linear Multivariate (LMUL) algorithm. These three multidimensional interpolation methods were applied to model the cancellous bone of the ankle. Overall, the RBAS algorithm performs the best, and the NNEI method can be a backup in case of large-scale computation. The LMUL algorithm becomes problematic when the query points are outside the bounding box.
