| Accurate computational simulations are crucial to understanding the causes and potential treatment of the 27,000 cerebral aneurysms that rupture annually in the United States. While the accuracy and number of computational simulations has increased, modeling and meshing the geometry, and providing proper boundary and loading conditions remains a complex process. It is because of these complexities that most simulations rely on a uniform wall thickness with constant, isotropic material properties. Unfortunately, the stresses calculated for a cerebral aneurysm are highly susceptible to the quality of the model being used. This thesis addresses the issue of model fidelity and mesh quality by presenting algorithms to determine an approximate, or equivalent, wall thickness for patient-specific cerebral aneurysms and to create a new boundary-layer mesh sizing-function. The equivalent wall thickness is found by deforming the mesh of a healthy blood-vessel onto the aneurysm model, where the stretching of the mesh elements represents the weakening blood-vessel wall. A medial-ball based sizing function prevents overlapping in boundary-layer meshes and increases the usability of anisotropic, prismatic elements. With a more exact model of the aneurysm wall, clinicians will be provided multiphasic diagnostic, preventative, and curative information. Furthermore, future simulations will allow for early diagnosis and quantification of potential treatment options; with early intervention culminating in the best course of action being chosen for each individual patient. The contributions of this thesis are demonstrated through a series of fluid-structure interaction simulations, where use of the equivalent wall-thickness is shown to have a significant effect on the wall stresses. Furthermore, the deformation process used to calculate the equivalent wall thickness is shown to be able to map the anisotropic material directions of the vessel tissue onto the aneurysm itself. |
