Development and validation of patient-specific finite element models of patients with total hip arthroplasty : towards a clinical tool

Abstract

The development and validation of patient-specific finite element (FE) models of the pelvis of patients who have had total hip arthroplasty (THA) is presented. The finite element method (FEM) has been used widely in orthopaedic biomechanics for various research purposes, but the focus of this project is to develop it one step further as a clinical tool. Hence an automatic mesh generation method that can create FE meshes from patient CT data sets has been developed. In doing so, a special attention has been paid to cases where patient CT data sets do not cover the whole bone of interest, which is not uncommon in clinical practice. Thus a hybrid method that can generate FE models from incomplete/sparse patient data sets supplemented from the Visible Human generic database has been developed. Material properties were obtained also from patient CT scans to be assigned to the FE mesh and a novel approach to assign material properties to large FE elements generated from sparse data sets has been developed as well. Using these models, mechanical simulations testing various hypothesis have been conducted. The performance of our FE model has been validated in two ways - (1) sensitivity analysis which examines the stability of the model's numerical performance; (2) mechanical testing with real pelves to compare predicted values from the model and actual results from the experiment. Both tests have shown that our model is capable of producing reproducible results that are accurate and meaningful. Two applications of our method are presented. Firstly the changes in load transfer pattern before and after total hip arthroplasty have been investigated with our FE models. Our results indicate that some areas around the acetabulum might be prone to stress shielding after the prosthetic cup has been inserted. We also applied our model to clinical CT scans from a long term follow-up study of patients with THA. The mechanical simulation results complement the previous densitometry study performed on the same CT scans and give more insight to the remodelling occurring after THA. These show the possible areas where our method can be applied. Future works also include applying our model to various other clinical cases such as fractured hip and modelling cysts in patients with THA. The thesis ends with discussions about practical issues with our model and its limitations.