Date of Award
Doctor of Philosophy in Chemical Engineering - (Ph.D.)
Chemical and Materials Engineering
Roman S. Voronov
Piero M. Armenante
Linda Jane Cummings
The challenges involved in modeling biological systems are significant and push the boundaries of conventional modeling. This is because biological systems are distinctly complex, and their emergent properties are results of the interplay of numerous components/processes. Unfortunately, conventional modeling approaches are often limited by their inability to capture all these complexities. By using in vivo data derived from biomedical imaging, image-based modeling is able to overcome this limitation.
In this work, a combination of imaging data with the Lattice-Boltzmann Method for computational fluid dynamics (CFD) is applied to tissue engineering and thrombogenesis. Using this approach, some of the unanswered questions in both application areas are resolved.
In the first application, numerical differences between two types of boundary conditions: “wall boundary condition” (WBC) and “periodic boundary condition” (PBC), which are commonly utilized for approximating shear stresses in tissue engineering scaffold simulations is investigated. Surface stresses in 3D scaffold reconstructions, obtained from high resolution microcomputed tomography images are calculated for both boundary condition types and compared with the actual whole scaffold values via image-based CFD simulations. It is found that, both boundary conditions follow the same spatial surface stress patterns as the whole scaffold simulations. However, they under-predict the absolute stress values approximately by a factor of two. Moreover, it is found that the error grows with higher scaffold porosity. Additionally, it is found that the PBC always resulted in a lower error than the WBC.
In a second tissue engineering study, the dependence of culture time on the distribution and magnitude of fluid shear in tissue scaffolds cultured under flow perfusion is investigated. In the study, constructs are destructively evaluated with assays for cellularity and calcium deposition, imaged using µCT and reconstructed for CFD simulations. It is found that both the shear stress distributions within scaffolds consistently increase with culture time and correlate with increasing levels of mineralized tissues within the scaffold constructs as seen in calcium deposition data and µCT reconstructions.
In the thrombogenesis application, detailed analysis of time lapse microscopy images showing yielding of thrombi in live mouse microvasculature is performed. Using these images, image-based CFD modeling is performed to calculate the fluid-induced shear stresses imposed on the thrombi’s surfaces by the surrounding blood flow. From the results, estimates of the yield stress (A critical parameter for quantifying the extent to which thrombi material can resist deformation and breakage) are obtained for different blood vessels. Further, it is shown that the yielding observed in thrombi occurs mostly in the outer shell region while the inner core remains intact. This suggests that the core material is different from the shell. To that end, we propose an alternative mechanism of thrombogenesis which could help explain this difference.
Overall, the findings from this work reveal that image-based modeling is a versatile approach which can be applied to different biomedical application areas while overcoming the difficulties associated with conventional modeling.
Kadri, Olufemi E., "Overcoming conventional modeling limitations using image- driven lattice-boltzmann method simulations for biophysical applications" (2018). Dissertations. 1388.