Effect of Tissue Structure (and) Disease on Simulated Arrhythmias in the Human Heart

Ventricular Fibrillation (VF) is a severe cardiac arrhythmia. Early experiments provided evidence that the mechanism of VF is consistent with re-entry. In 3D the sources of re-entrant waves are lines of phase singularity called ‘filaments’. Filament interactions and filament numbers can be used to quantify the complexity of activation patterns in simulated VF. The aim of this thesis is to study the effect of tissue structure, shape, initial conditions, and region of scar on filament dynamics using computational modelling. Transmural heterogeneity in 3D slab tissue representing the ventricular wall did not show important difference in the number of filaments. Configuration of filaments were influenced by transmural heterogeneity. With transmural heterogeneity, clustering of filaments were observed near slow conducting border resulting in increase of filament life time. To study the effect on the shape of the tissue on filaments, filament dynamics in 3D slab tissue was compared with filament dynamics in an idealized human left ventricle (LV) with similar apex base dimension and wall thickness. The volume of idealized LV is about twice the volume of 3D slab. Results showed idealized LV had twice the number of filaments compared to slab geometry especially with steeper restitution dynamics. This thesis highlights non-linear behaviour of activation patterns during VF in 3D where small changes had a large influence on the number of filaments in both 3D slab and idealized LV especially with steeper restitution dynamics. Pre-existing scar tissue in VF patients can act as a source of anatomical re-entry and pin re-entrant filaments to the scar boundary. However, the interaction of scar with complex activation during VF is not well understood. This thesis investigated how simulated scars (circumferential, transmural and sub-endo transmural) with varying size and with either regular or with irregular scar boundary, influenced re-entrant filaments in simplified computational models of 3D slab and idealized LV. Circumferential scar did not show any influence on filament dynamics and clustering of filaments to scar boundary. Increased radius of transmural and sub-endocardial transmural scar had more clustering of filaments to the scar boundary especially with steeper restitution dynamics. Sub- endocardial transmural scar with increased radius had more clustering of filaments compared to transmural scar. Region of scar with irregular boundary had more clustering compared to scar with regular boundary. Generally transmural scar and sub-endocardial transmural scar with regular or irregular boundary did not increase the number of filaments much but increased the length and lifetime of filaments due to clustering of filaments to the scar boundary. Since filaments pin to the boundary of the scar region, it can be hard to remove them using defibrillation techniques. It might be necessary to measure the radius and the depth of pre-existing scar tissue accurately in VF patients, as the bigger scar is likely to have more complex activation in VF. In this way it might be easier to predict the strength of defibrillation to use, in order to stop VF.