Patient-specific parameterization of a left-ventricular model of cardiac electrophysiology using electrocardiographic recordings

Abstract

Introduction: Image-based models of human electrophysiology (EP) are increasingly considered as a clinical research tool. However, current clinical EP models are typically not patient-specific as they mostly rely on generic data or are computationally not efficient enough to fit with clinical time scales. Objectives: This study aimed to develop an efficient, clinically-compatible automated workflow for patient-specific parameterization of cardiac EP models using non-invasive standard ECG recordings. Specifically, we focused on the parameterization of the depolarization phase during sinus activation to reproduce QRS morphology. Methods: Two MRI-based left-ventricular (LV) models, A and B, were utilized. A simplified activation model was defined based on the assumption that activation patterns are determined by the locations, x, of septal, anterior and posterior fascicle of the His-Purkinje system (HPS), their relative activation timings,t, and the conduction velocities, v, within LV wall and HPS. A reaction-eikonal model was employed to compute activation sequences, source distributions and ECGs. Latin hypercube sampling was used to sweep the input parameter space [x, t, v]. Quantitative comparison between QRS complexes of simulated and measured ECGs was performed using a normalized correlation coefficient and L2 norm. Results: Activation sequences with corresponding QRS complex were simulated in approximately 25 seconds. Inherent morphological characteristics of the QRS complex could be represented by our model parameter space [x, t, v]. Correlation coefficients and L2 norms of 0.86 and 20.22 were attained for model A, and 0.93 and 3.06 for model B, respectively. Discussion: The feasibility of generating patient-specific LV activation sequences based on measured QRS complexes in non-invasive ECG recordings was demonstrated. The efficiency of the proposed model will facilitate its use in a future more general framework for data-driven clinical EP model parameterization.