In this project, we aim at improving a model devised to simulate the generation and transmission of electrical impulses by cardiac cells. This model is used to study the mechanisms of initiation and sustainment of vortices of excitation waves which underlie cardiac tachyarrhythmias. In a subsequent phase, and if the model can realistically reproduce vortices of excitation, we will study how membrane functions can be altered to terminate such vortices. We have developed a parallel version of our code using MPI. Based on our performance tests, using 32 processors on the IBM SP, it will take 10 hours (wall clock) to obtain an episode of tachycardia; i.e. 1 sec of vortex-like activity, in a small piece of cardiac tissue of 2 cm x 2 cm. The same simulation requires a little more than 8 days on a SPARC 10 model 512. During the period of support, we will improve the matrix inversion procedure of the current version. We believe that it is possible to divide the computation time by a factor ranging between 3 and 5 by using a preconditioning procedure suitable for parallelization
Showing the most recent 10 out of 20 publications