High-fidelity model of the human heart: An immersed boundary implementation

Abstract

We present a multi--physics computational model of the human heart accounting for the electrophysiology, elasto-mechanics, and hemodynamics, including their complex interactions. The model is accurate and computationally efficient and, thanks to the implementation on GPU architectures, it allows cardiovascular simulations of physiologic and pathologic configurations within a time--to--solution compatible with clinical practice. Results are shown for healthy conditions and for myocardial infarction with the aim of assessing the reliability and predictive capabilities of the model which can be used to anticipate the outcome of surgical procedures or support clinical decisions.