Premium
A massively parallel computational electrophysiology model of the heart
Author(s) -
Vázquez M.,
Arís R.,
Houzeaux G.,
Aubry R.,
Villar P.,
GarciaBarnés J.,
Gil D.,
Carreras F.
Publication year - 2011
Publication title -
international journal for numerical methods in biomedical engineering
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.741
H-Index - 63
eISSN - 2040-7947
pISSN - 2040-7939
DOI - 10.1002/cnm.1443
Subject(s) - massively parallel , polygon mesh , computer science , scalability , computational science , key (lock) , cardiac electrophysiology , transient (computer programming) , computational model , ideal (ethics) , computational resource , parallel computing , scale (ratio) , computational complexity theory , theoretical computer science , distributed computing , simulation , algorithm , electrophysiology , computer graphics (images) , medicine , philosophy , computer security , physics , epistemology , quantum mechanics , database , operating system
Abstract This paper presents a patient‐sensitive simulation strategy capable of using the most efficient way the high‐performance computational resources. The proposed strategy directly involves three different players: Computational Mechanics Scientists (CMS), Image Processing Scientists and Cardiologists, each one mastering its own expertise area within the project. This paper describes the general integrative scheme but focusing on the CMS side presents a massively parallel implementation of computational electrophysiology applied to cardiac tissue simulation. The paper covers different angles of the computational problem: equations, numerical issues, the algorithm and parallel implementation. The proposed methodology is illustrated with numerical simulations testing all the different possibilities, ranging from small domains up to very large ones. A key issue is the almost ideal scalability not only for large and complex problems but also for medium‐size meshes. The explicit formulation is particularly well suited for solving this highly transient problems, with very short time‐scale. Copyright © 2011 John Wiley & Sons, Ltd.