Patient-specific blood flow modelling

Cardiovascular diseases that affect arteries often result in vessel blockage which can have dramatic consequences on the patient's life. In situation of obstructed arteries, graft surgery is a common treatment to improve blood flow circulation. Nevertheless, in the particular case of lower limb arteries, the rate of success of this medical intervention is quite low. The patients suffer from graft blockages induced by Intimal Hyperplasia (IH). Local flow patterns are known to have a strong influence on the development of IH. In order to better understand the causes of graft failure, Computational Fluid Dynamics is used to study the role of hemodynamics in each patient bypass geometry. This thesis focuses on both the automatic mesh generation from medical images and the blood flow simulation. First, the patient graft shape is constructed from image segmentation with particular focus on the distal junction with the native artery. The initial mesh is then adapted from the surface curvature to take into account the influence of the geometrical specificities. An anisotropic metric field is generated on the surface and in the volume in order to reduce the total number of degrees of freedom. The obtained mesh is considered more suitable for numerical computation. Secondly, fluid simulations were run using experimental data from each patient to set the boundary conditions on the model. The localization of hemodynamic indicators allows to determine areas of the geometry most susceptible to the development of strong cell proliferation. Realistic geometries are shown to have a strong influence on the flow patterns and therefore on the pathology progression.