The Impact of Boundary Conditions in Patient-Specific Coronary Blood Flow Simulation

In patient-specific coronary blood flow modelling, the computational domain is defined as a 3D mesh of the major coronary arteries reconstructed from medical images. The blood flow itself is described through the flow equations that are defined over this computational domain and further solved by methods of computation fluid dynamics (CFD) in terms of pressure and velocity fields. One of the major factors determining the accuracy of the computed flow fields is the appropriate specification of flow conditions on the boundaries of the domain that represent the impact of the remaining (undefined) parts of the cardiovascular system (CVS). The boundary conditions (BC) can be assigned either explicitly or implicitly; the difference being in the presence, or not of interaction between the domain (3D arteries) and the upstream/downstream vasculature. Application of more advanced implicit BC provides the means to model the adjustment of blood flow depending on the downstream impedance and intramyocardial pressure, which results in "more realistic" computed blood fields. However, this also introduces a level of uncertainty, as the blood flow in the required zones of CVS is unknown a priori and generally cannot be measured in clinical practice due to the highly invasive nature of the procedure. This paper presents an analysis of the requirements for boundary conditions in the context of physiologically realistic simulations of coronary blood flow. This will be elucidated through the analysis of 3D blood flow in patient-specific left coronary artery (LCA). Moreover, a comparison between implicit and explicit BC and the methods of modelling patient-specific flow within the rest of CVS will be provided.