Bayesian Calibration of Electrophysiology Models Using Restitution Curve Emulators.

Sam Coveney,Richard D Wilkinson,Richard H Clayton,Jeremy E Oakley,Steven A Niederer,Cesare Corrado

doi:10.3389/fphys.2021.693015

Sam Coveney, Richard D Wilkinson + Show 4 more

Open Access

https://doi.org/10.3389/fphys.2021.693015

Copy DOI

Abstract

Calibration of cardiac electrophysiology models is a fundamental aspect of model personalization for predicting the outcomes of cardiac therapies, simulation testing of device performance for a range of phenotypes, and for fundamental research into cardiac function. Restitution curves provide information on tissue function and can be measured using clinically feasible measurement protocols. We introduce novel “restitution curve emulators” as probabilistic models for performing model exploration, sensitivity analysis, and Bayesian calibration to noisy data. These emulators are built by decomposing restitution curves using principal component analysis and modeling the resulting coordinates with respect to model parameters using Gaussian processes. Restitution curve emulators can be used to study parameter identifiability via sensitivity analysis of restitution curve components and rapid inference of the posterior distribution of model parameters given noisy measurements. Posterior uncertainty about parameters is critical for making predictions from calibrated models, since many parameter settings can be consistent with measured data and yet produce very different model behaviors under conditions not effectively probed by the measurement protocols. Restitution curve emulators are therefore promising probabilistic tools for calibrating electrophysiology models.

Highlights

Cardiac electrophysiology models reconstruct electrical activation of the heart at cell, tissue, and organ scale
We have demonstrated a way to emulate restitution curves by using Gaussian processes to predict the principal component coordinates of restitution curves from model parameters
We developed an analytical likelihood function for effective refractory period (ERP) observations, which is especially useful for calibration with Restitution Curve Emulators (RCEs)

Summary

Introduction

Cardiac electrophysiology models reconstruct electrical activation of the heart at cell, tissue, and organ scale. Alongside biophysically detailed models of cardiac cellular electrophysiology, Restitution Curve Emulators phenomenological models have been developed that capture action potential shape and rate dependence without an explicit representation of ion channel behavior (Fenton and Karma, 1998; Mitchell and Schaeffer, 2003; Corrado and Niederer, 2016). These models have fewer parameters than more detailed models and can be solved relatively quickly, but the association between model parameters and emergent properties remains complex (Fenton et al, 2002). Clinical data are typically noisy and sparse so recent developments have included a set of approaches that take into account uncertainties in the data to create probabilistic models (Konukoglu et al, 2011; Coveney et al, 2020; Dhamala et al, 2020), as well as new models designed with uncertainty in mind (Pathmanathan et al, 2019)

Methods

Results

Conclusion