Building a heart model from cells
Christopher Oehmen, University of Memphis
Computational models exist for electrically active single cells of almost every type found in the heart. While these cell models have offered insight into mechanisms of electrical activity in normal, drug applied, and pathophysiological states, there is much information to be gained from understanding the interaction of these cells when they are electrically joined in a functional myocardial tissue. Because of computational complexity, several theories have been put forth to simplify the modeling of cardiac tissues by neglecting the details relevant on the cellular level. With the constantly improving computational facilities and software available to the investigator, and the growing body of knowledge concerning genetic expression in cardiac systems, the integration of models of cardiac function over many orders of spatial and temporal magnitude has become a reality. Tissue or organ models built-up from cellular models offer the advantage that they can incorporate information from heterologous gene expression experiments to explain pathological events at the tissue organ level. We have laid the foundation for exploring tisues using single cell models for heterogeneous tissue structures (such as endocardium and epicardium) by developing a computational platform which can incorporate existing computational models into a tissue-level structure. Our results show that (1) incorporating heterogeneous cell types into our tissue data structure and using CVODE (Lawrence Livermore National Laboratory) to integrate the resulting state variables and cell-cell interactions simultaneously results in tractable solution times on a single processor machine for small collections of cells; and (2) one role of heterogeneity of cell types in the rabbit sinoatrial node (our prototype implementation) is to provide protective mechanisms for maintaining cardiac automaticity. It is hoped that porting our platform to a massively parallel environment will provide a foundation for future studies which will continue to integrate protein-level information to the level of whole heart behavior.
Abstract Author(s): Christopher Oehmen, Semahat S. Demir, Ph.D