Approximation of Patient-Specific Blood Flow in Arterial Bifurcationss

Obtaining detailed, patient-specific blood flow information is very useful in detecting and monitoring cardio-vascular diseases. Nevertheless, available clinical data are usually poorly resolved and contain a signifcant amount of noise. Therefore, current approaches rely on computational fluid dynamics to calculate high quality flow data. However, this method is hardly usable in the daily clinical routine due to the required technical supervision and the huge computational resources needed for the numerical simulation. At the same time, only a very limited amount of available flow imaging data are used. The talk will discuss an interpolation approach based on proper orthogonal decomposition, allowing the approximation of patient-specific flow fields based on a pre-computed database. We will demonstrate that this method is able to dramatically improve the quality of the direct measurements, in particular indirect, flow-derived data such as wall shear stress. Additionally, unmeasurable quantities like pressure can be estimated, providing a practical and clinically feasible alternative to fully-blown, time-consuming, patient-specific flow simulations.