Multiscale Computational Modeling of Heterogeneous Stem-Derived Cardiac Tissues

The interpretation of electrophysiological measurements in in vitro cardiac models raises fundamental mathematical and computational questions. In particular, multielectrode array (MEA) recordings provide only indirect access to the underlying cellular dynamics, and their quantitative meaning depends critically on the continuum models and numerical discretizations used to describe the tissue-electrode system.

In this seminar, I present a multiscale computational framework based on the bidomain theory for 2D heterogeneous derived cardiac tissues, extended to account for electrode-tissue coupling. We compare three modeling paradigms: the classical bidomain model, a formulation with electrode- specific boundary conditions, and a fully coupled MEA model including electrode feedback. Although grounded in the same biophysical principles, these approaches lead to distinct variational formulations and algebraic structures, with relevant implications for stability, accuracy, and computational scalability.
Building on this setting, I discuss a large-scale in-silico correlation study between field potential (FP) biomarkers and action potential (AP) features in tissues composed of atrial-like and ventricular-like phenotypes. By combining bidomain simulations with a 256-electrode MEA representation, we quantify how spatial sampling and tissue heterogeneity affect FP-AP correlations, identifying the FP metrics that most reliably reflect AP properties.

Overall, the proposed framework provides a predictive and mechanistic bridge between cellular electrophysiology and experimentally accessible MEA readouts, supporting a more rigorous interpretation of cardiac assays based on stem derived cardiomyocytes.


Contatto:
andrea1.manzoni@polimi.it