Cardiac fibrosis is widely recognized as a key contributor to arrhythmogenesis, yet its representation in computational models remains highly variable. Existing approaches differ substantially in how structural and electrophysiological remodeling are incorporated, but assessment of modelling choices upon simulated behaviors is largely lacking.
This work explores a novel framework for representing fibrotic remodeling based on stochastic partial differential equations, aimed at capturing spatial heterogeneity in a flexible and scalable manner. The proposed approach is first investigated in two-dimensional tissue models, where fibrosis patterns can be generated in a controlled manner. Electrophysiological simulations are performed under varying assumptions on conduction and ionic properties to examine how different representations of fibrosis influence activation dynamics, electrogram morphology, and fractionation. These initial studies are used to assess the model's behavior and guide further development. The framework is further being adapted to volumetric anatomical twin models, in which fibrotic remodeling can be incorporated with varying spatial patterns.
Within this setting, we consider how different fibrosis configurations may influence electrophysiological propagation and arrhythmia inducibility. Simulations incorporating controlled activation sequences as pre-pacing maps are being explored to probe these interactions in a systematic manner. In parallel, we reflect on how this approach relates to commonly used fibrosis modeling strategies, with a focus on the implications of different modeling assumptions.
This work aims to develop a universal algorithm able to reproduce and characterize different types of fibrotic patterns in cardiac simulations and investigate how different representations might influence electrophysiological outcomes.