Towards predictive computer simulations in cardiology: Finite element analysis of personalized heart models

Abstract

AbstractThe goal of this manuscript is to demonstrate the feasibility of our recent numerical developments towards predictive computer simulations in cardiology. In contrast to existing weakly coupled and strongly coupled monolithic approaches in the literature, we utilize a fully implicit staggered solution scheme that enables to study the strong excitation‐contraction coupling of the heart tissue in the monodomain and the bidomain setting through finite element simulations. On the constitutive level, we employ the recently proposed modified Hill model (CMAME 315; 434‐466, 2017) that treats the myocardium as an electro‐visco‐active material. All the time integrations are evaluated via an implicit backward Euler scheme that ensures unconditional stability. The performance of the framework is demonstrated by means of two clinically relevant and interesting examples. Firstly, basic deformation characteristics of a personalized left ventricle model such as rotation, twist and longitudinal shortening are simulated along with a physiological pressure‐volume relation. In addition, the results of the viscoelastic and elastic solutions are compared. In the second example, we simulate a continuously beating virtual biventricle model having dyssynchrony and imitate two different cardiac resynchronization therapy attempts in order to improve the cardiac output. The corresponding electrocardiograms and left ventricle volume‐time relations are recorded during the simulation and compared to the healthy case.