Image-based CFD simulations have been proposed to investigate the local hemodynamics inside human cerebral aneurysms. The knowledge of the computed 3D flow fields can be used to assist clinical risk assessment and treatment decision making. To this end, it is desired to know the reliability of CFD for cerebral blood flow simulation, and then to provide clinical feedbacks. In this thesis work, a simulation pipeline is built up and the accuracy of CFD is quantitatively evaluated via virtual angiography method.
The proposed simulation pipeline is first tested on in-vitro phantom cases under fully controlled conditions. A clinical scenario with stent placement is also considered and tested on a simplified phantom. After that, the simulation pipeline is extended to the in-vivo patient cases where several uncertainties are unavoidable. In general, close similarities are found in terms of spatial and temporal variations of CA distribution between the experimentally or clinically acquired and the virtually constructed angiograms, proving CFD has ability to reproduce the cerebral blood with reasonable accuracy. The results also demonstrate that the input flow BCs can place influence on the computed flow ields. In this work, the reliability of the CFD for simulating cerebral blood flow is well confirmed. This indicates that CFD has the potential to be applied in the clinics and provide useful feedbacks with confidence.