Cardiac arrhythmias such as atrial fibrillation (AF), ventricular fibrillation (VF) and ventricular tachycardia (VT) are growing causes of morbidity and mortality across the world. To aid the discovery of the mechanisms driving these arrhythmias, in-silico models have been developed to simulate signal propagation in cardiac tissues.
Continuum models such as monodomain and bidomain approaches are the most common methods to represent multicellular electrical activity. These approaches have been successfully applied on the whole heart and large-scale tissues with acceptable approximation. However, they may not be appropriate for microscopic scale simulations. It is known that cellular remodelling, including the development of scars, changes in ion channels or gap junctions play a role in arrhythmogenesis. However, in patients and laboratory models, where these occur simultaneously, the direct effect and importance of single factors are difficult to determine.
Discrete-cell models represent action potential generation and propagation in individual cells and provide a more accurate cell-level simulation, but at much greater computational cost.
In this study, 2D simulations of ventricular tissues with a range of gap junction distributions, informed by biological staining experiments, are performed. Both continuum and discrete-cell models are applied to the same tissues and their propagation patterns are compared. While the continuum model accurately captures propagation with uniform Cx43 distribution, the discrete-cell model provides better accuracy with heterogeneous distributions.