A Methodology to Improve Human Ventricular Models for the Investigation of Cardiac Arrhythmias

Abstract

Introduction: A number of human action potential (AP) models can be found in the literature for the study of ventricular arrhythmias. One of the latest proposed models is the Carro-Rodriguez-Laguna-Pueyo (CRLP). Although this model reproduces many human ventricular physiological features, it lacks the characteristic AP peak and dome. In this work we propose a modification of the L-type Calcium current (ICaL) to improve its AP shape and to best fit the model to experimental results.Methods: Both the time constants and the steady-state value of the ICaL inactivation gates were modified, with validation performed at two different levels: 1)The results for the new ICaL formulation were compared against three different experimental datasets using in silico simulations of the experimental protocols. 2)A set of arrhythmic risk markers were calculated using the modified AP model, including: systolic and diastolic intracellular calcium concentration ([Ca2+]i), AP duration, AP triangulation at different pacing frequencies, and APD adaptation to abrupt cycle length changes.Results: The mean square error of steady-state ICaL inactivation was 25% lower than in the original CRLP model and satisfactorily reproduced the available experimental data. The modified time constants of the ICaL inactivation gates led to physiological AP peak and dome shapes. All the computed arrhythmic risk markers remained in the physiological range or were very similar to the ones reported for the original CRLP model.Conclusions: The proposed modification in the CRLP model improves the behavior of the ICaL current and the AP shape, providing a model suitable for the investigation of ventricular arrhythmias. The strategy proposed in this work can be extended to other human ventricular cell models available in the literature.