Cardiac Electro-Mechanics: From CellML to the Whole Heart
My PhD thesis in Bioengineering at the University of Auckland, completed March 2005. Supervised by Prof. Peter Hunter.
We have developed a computational modelling and simulation framework for cardiac electromechanics which for the first time tightly couples cellular, tissue, and whole heart modelling paradigms. Application of the framework has been demonstrated in simulations of the electrical activation and mechanical contraction of cardiac myocytes, myocardial tissue, and models of ventricular structure and function.
The framework has been implemented as part of the CMISS computational modelling environment. This allows for detailed specification of tissue microstructure and the variation of cellular models and their material parameters over a geometric domain of interest. Through previous work in CMISS we have access to a large pool of existing models and methods on which to base this work.
A key aspect of our framework is the implementation of methods allowing the dynamic specification of mathematical models at run-time using CellML – an XML language designed to store and exchange computer based biological models. We use these methods to enable the specification of cellular level models using CellML, thus providing a very powerful method to simply “plug” cellular models into the distributed tissue and organ level models. While in this work we have developed specific cellular models, and their CellML description, the framework is now developed to the point where a non-expert user could pull out a given model and plug-in their own model. This allows cellular modellers to develop their models and then bring them into a distributed model of tissue or organ physiology.
We present a review of cellular electrophysiology and mechanics models. As a demonstration of the ability of CellML to describe cardiac cellular models in a computationally useful manner we have encoded the majority of these models into CellML. The CellML descriptions of these models are provided as they were used in the computational simulations which generated all results presented for the models directly from the CellML encoded models.
While the work described in this thesis focuses on cardiac electro-mechanics, the framework has much wider applicability. As such, this work forms the beginning of a generalised simulation framework for the IUPS Physiome Project, with the goal being models describing entire organisms from proteins through to organ systems.