Key result
New fitting pipeline yields surrogate models matching realistic human electrophysiology models in coherence and fractionation.
Why the study?
Existing cardiac electrophysiology models often fit only action potential morphology or single restitution curves, leading to overfitting and failure to ensure appropriate physiological and dynamical behavior.
A novel fitting pipeline for cardiac electrophysiology models successfully reproduces complex tissue-level dynamics with fewer parameters, enabling faster simulations for digital twin applications.
In the last 30 years, people have devoted great efforts to generate models that reproduce the electrophysiological complexity of patient hearts. These models are referred to as digital twins and require a large number of parameters to be calibrated through experimental data and careful analysis. The choice of observables to perform the fit is crucial in determining if the model will provide useful information. In the literature, it is common to find articles in which the cardiac action potential (AP) morphology or a single restitution curve serves as the only data source for the fit. This problem leads to a situation in which the model reproduces a limited set of features and is susceptible to overfitting, which means that there is no way to ensure that the model behaves appropriately in a physiological and dynamical way. In this work, we present a new fitting pipeline that incorporates AP duration, conduction velocity, and activation time restitution features in single-cell and tissue. These observables are commonly measured in experimental and clinical setups and provide information at a tissue level; thus avoiding the need for slow and convoluted experiments or sampling them from other systems. Furthermore, the model we fit is a phenomenological model with a low number of parameters; hence, we are able to create a one-to-one map between the observables and most of the parameters. Our results show that the pipeline is able to reproduce important characteristics of the restitution relations (Fig. 1) and the dynamical complexity of realistic human models in single-cell and tissue (Fig. 2). In Fig. 2 we show the comparison of the time evolution of the AP excitations of 8 realistic human electrophysiology models (3 of them approved by the Food and Drug Administration for proarrhythmic risk assessment [1]) to our surrogate fits. Our results show that the fitting pipeline produces surrogate models with the same level of pattern coherence and fractionation as their patient model counterparts. Moreover, the simple structure of the surrogate models allows us to increase the computational speed of the simulations, bringing us one step closer to a practical real-time digital twin solution. Moreover, we show that our method can be applied to ventricle and atria systems, corroborating the universality of our new fitting paradigm. Finally, we discuss the clinical applicability of the pipeline and suggest optimizations to it.
No takes yet. Share an insight, caveat, or question.
H Velasco-Perez (2026) studied Cardiac electrophysiology modeling. New fitting pipeline incorporating AP duration, conduction velocity, and activation time restitution features vs. 8 realistic human electrophysiology models was evaluated on Reproduction of restitution relations and dynamical complexity. A new fitting pipeline successfully produced surrogate models with the same level of pattern coherence and fractionation as 8 realistic human electrophysiology models.
Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context: