Development and Validation of a Voxel Based Element by Element Subject-Specific Finite Element Modelling of the Knee
Osteoarthritis (OA) is a painful, degenerative joint disease marked by mechanical and morphological alterations. It affects 1 in 8 Canadians, half of which present it at the knee. Quantitative computed tomography-based finite element (QCT-FE) modelling is a non-invasive in vivo research tool which can be used to study the mechanical behavior of bone in OA. The main drawbacks of this approach are due to its need for large computing power. The time needed to run a model is in the order of hours to a day due to the many steps involved, specifically meshing and material mapping, making it unfeasible as a research tool for a large data set or in a clinical setting. The objectives of this study were to 1)develop and validate a faster, element-by-element (EBE) FE modelling approach with a voxelized mesh, 2)assess the repeatability of the new approach and study whether mechanical metrics can discriminate people with OA, and 3)study links between mechanical metrics and OA-related knee pain. We developed EBE QCT-FE methodology that solves a model within 10 minutes. This approach was validated against the gold-standard approach (QCT-FE), outputting similar results: R2=0.97 and root mean squared error (RMSE%) =1.85%. We found that von Mises stress had low precision errors (CV%RMS=9.7%) and values 141% higher medially and 65% higher laterally in individuals with OA (n=7) when compared to healthy controls (n=7). Individuals with OA and severe nocturnal pain (n=9) presented 26-135% higher lateral stress and 24-48% lower medial stress compared to individuals with OA and without nocturnal pain (n=17). With the same dataset, we also found that individuals with severe nocturnal pain had 75%-116% higher medial strain and 45% lower lateral strain compared to individuals with no nocturnal pain. Similarly, the bone in individuals with severe pain was 77-117% closer to failure medially and 44% safer laterally compared to individuals with no nocturnal pain. With this study, we developed a fast, repeatable, and valid approach to generate FE models of the knee. This approach can be used to further study the mechanical behavior of bone in OA progression and pain pathogenesis in larger data sets. Preliminary results suggested that loading patterns differ between people with and without OA, and that mechanical metrics may be related to pain in OA.
Finite Element, Osteoarthritis, Knee, Quantitative Computed Tomography, Pain, Stress, Strain
Master of Science (M.Sc.)