Development of an Anatomically Realistic Left Ventricle Physical Model and Multi-Modality Experimental Platform for Validation of Patient-Specific Computational Simulations

Recent trends in bioengineering, also supported by the FDA [1], highlight the importance of experimental validation of numerical solvers used in medicine in order to use numerical solvers for surgical planning. Efforts to improve the diagnosis of left heart diseases have pointed to the importance of hemodynamic patterns in the left ventricle [2] and the use of CFD simulations could aid in repair and treatment of left heart disease. In this study, we aim to experimentally validate the Curvilinear Immersed Boundary solver (CURVIB) [3] to use patient specific data for simulations. As a first step, an idealized left heart model with a single deforming wall was used for comparison of the diastolic intra-ventricular flow field between experimental and CFD results. The inputs for the numerical solver include the dynamics of LV wall motion as well as the mitral and aortic flows. We have achieved good agreement between the experimental and CFD and our goal is to translate the solver to use clinical data. We present a reconstruction method for the LV deformation from normal volunteer MRI images and an anatomically realistic LV physical model that has been designed for validation.