| dc.description.abstract | Metastatic cancer, the spread of cancer cells from the primary cancer site, is responsible for many cancer-related complications and deaths. Particularly, metastatic cancer affecting the vertebra leads to the degradation of bone quality and architecture, which weakens the vertebra's load carrying capacity and puts the patient at high risk of skeletal adverse events (SAE), such as pathological vertebral fractures (PVF) and spinal cord compression. Therefore, there remains a strong need for the accurate assessment of fracture risk of metastatic vertebrae.
In this thesis, we propose a comprehensive computational framework for the modeling and validation of the deformation and failure response of human metastatic vertebrae. First, we develop an image-based pipeline for the generation of patient-specific finite elements (FE) models starting from CT images of human metastatic vertebrae. We adopt a viscoelastic, viscoplastic cortical bone modelfor the elastic and plastic response of bone and a damage model for the softening of bone. We utilize the SUMMIT computational solid mechanics framework to perform large-scale, parallel simulations of these vertebral models under compression loading.
We validate the computational framework against an experimental dataset of 10 metastatic human vertebrae, consisting of 1) CT image data and 2) experimental load-displacement curves obtained from uniaxial compression testing,provided by researchers at the University of Bern. Using our proposed pipeline, we generate homogenized finite element (hFE) models from the obtained CT images and calibrate the material model parameters by solving the boundary value problem for a selected model using trial values of the material model parameters until we obtain a simulated load-displacement curve that approximately matched the corresponding experimental load-displacement curve. Then, we conduct simulations of the other 9 models and compare the simulated and experimental load-displacement curve to assess the validity of our models.
We show that the proposed approach provides quantitative predictions of the experimental stiffnesses and failure strengths of metastatic vertebrae with good accuracy regardless of the type of metastases the vertebrae exhibit. In addition, we show that by capturing the unique, spatially varying bone volume density in the vertebrae, we are able to obtain detailed descriptions of the local stress and damage responses. From this, we achieve a better understanding of the role metastases play in the deformation and damage response of metastatic vertebrae. | |
