Parallel Simulation of Scroll Wave Dynamics in the Human Heart Using the FEniCS Framework

Abstract

In the heart, scroll waves are three-dimensional self-sustaining spiral waves of electrical excitation. They arise only in pathology and underlie dangerous arrhythmias. Computer simulation of human cardiac arrhythmias is time-consuming and requires parallel computing. In addition, heart simulation is a complicated multilevel task (cell, tissue, organ); parallel multilevel simulation codes are difficult to create and maintain. Here we present an approach to parallel simulation of scroll waves in the left ventricle of the human heart utilizing the FEniCS framework, which allows developing software using near-mathematical notation and provides automatic parallelization. We used the Aliev-Panfilov cell model of heart electrical activity and studied the scalability of FEniCS implementation on parallel computing systems. © 2016 The Authors. Published by Elsevier B.V.