Parallelization of Three Dimensional Cardiac Simulation on GPU.

Abstract

<h4>Background</h4>The simulation of electrophysiological cardiac models plays an important role in facilitating the investigation of cardiac behavior under various conditions. However, these simulations often require a lot of computational resources.<h4>Methods</h4>To address this challenge, this study introduced a method for speeding up three-dimensional cardiac simulations using GPU parallelization. A series of optimizations was introduced, encompassing various aspects such as data storage, algorithmic enhancements, and data transfer.<h4>Results</h4>The experimental results reveal that the optimized GPU parallel simulations achieve an approximate 50-fold acceleration compared with their CPU serial program.<h4>Conclusion</h4>This investigation substantiates the considerable potential of GPUs in advancing the field of cardiac electrophysiology simulations.