Computational Fluid Dynamics Simulations of Mitral Paravalvular Leaks in Human Heart

Abstract

In recent years, computational fluid dynamics (CFD) has been extensively used in biomedical research on heart diseases due to its non-invasiveness and relative ease of use in predicting flow patterns inside the cardiovascular system. In this study, a modeling approach involving CFD simulations was employed to study hemodynamics inside the left ventricle (LV) of a human heart affected by a mitral paravalvular leak (PVL). A simplified LV geometry with four PVL variants that varied in shape and size was studied. Predicted blood flow parameters, mainly velocity and shear stress distributions, were used as indicators of how presence of PVLs correlates with risk and severity of hemolysis. The calculations performed in the study showed a high risk of hemolysis in all analyzed cases, with the maximum shear stress values considerably exceeding the safe level of 300 Pa. Results of our study indicated that there was no simple relationship between PVL geometry and the risk of hemolysis. Two factors that potentially played a role in hemolysis severity, namely erythrocyte exposure time and the volume of fluid in which shear stress exceeded a critical value, were not directly proportional to any of the characteristic geometrical parameters (shape, diameters, circumference, area, volume) of the PVL channel. Potential limitations of the proposed simplified approach of flow analysis are discussed, and possible modifications to increase the accuracy and plausibility of the results are presented.