Impact of tissue microstructure on a model of cardiac electromechanics based on MRI data

Abstract

Cardiac motion is a highly complex and integrated process of vital importance as it sustains the primary function of the heart, that is pumping blood. For this reason cardiac motion abnormalities are often associated with severe pathologies. Clinical non-invasive techniques can assess this fundamental connection between motion aberrant behaviour and pathology only to a certain extent. Computational models of cardiac motion would thus be of great help in linking local to global motion abnormalities and to pathology. No current model of the heart though is able yet to realistically simulate motion patterns in the healthy or diseased heart and therefore prediction of clinically relevant parameters such as ejection fraction, stroke volume, wall thickening and wall motion are not yet totally reliable.

It is well known that cardiac tissue microstructure, that is cardiac cells organisation into fibres and sheets, has an important role in cardiac motion, as suggested by experimental and computational research, but current models of cardiac function, although already embedding structural information in the models equations, are not yet able to correctly link structure to function and therefore predict realistic cardiac motion. The hypothesis underlying this project is that a more realistic representation of tissue structure within an electromechanical model of the heart, with fibre and sheet orientation extracted from data rather than mathematically defined, together with a more careful definition of tissue material properties, would better take into account the high heterogeneity of tissue structure, thus improving the predictive power of the model.

The aim of the project is to investigate how different settings of tissue structural arrangement affect the motion prediction of an electromechanical model applied to murine left ventricular geometries obtained from magnetic resonance imaging. The project relies on the integration of cardiac imaging data and mathematical modelling in the belief that realistic models of cardiac function need to reach the best compromise between the level of modelling detail and the amount and quality of information that can be actually obtained from data and used either to instruct or validate such models.