Subject Specific Finite Element Analysis of Bone: for evaluation of the healing of a leg lengthening and evaluation of femoral stem design

Abstract

The present thesis concerns employing the finite element method together with computed tomography to solve problems relevant to clinical orthopaedics.

The first part of the thesis describes a procedure of relating the gray-scale values from computed tomography to the local density of bone. Calibration procedures used today, simplify the composition of bone to consist of bone mineral and water-equivalent organic tissue. The procedure described in the present thesis accounts for both the mineral, collagen and fluid content in bone.

The geometry of the bone is extracted from the CT images to build the finite element models and assigned material properties according to local densities derived from the gray-scale values. Two different methods of building finite element models are described; voxel-based models, and geometry-based models.

Voxel-based finite element models are generated directly from the tomographic images and are thus relatively easy to use to study the stiffness of a bone segment. The voxel-based models are here used to assess the mechanical stability of a leg lengthening.

Geometry-based models are created by segmenting the tomographic images and building a 3D model, which in turn is meshed to create the finite element models. These models can be used to simulate surface conditions between implants and bone. They are here used to simulate the stress-shielding effect and stability of a cementless femoral stem in human cadaver femurs.