Electrophysiological model of the left ventricle: prediction of reentry circuits with fast simulations based on cellular automata applying clinical stimulation protocols

Abstract

Myocardial infarction (MI) is a common cardiovascular disease that causes irre- versible damage to the left ventricle (LV) myocardium, resulting in the formation of

scar tissue. When this phenomenon occurs, reentry circuits appear, generating alter- native conduction channels. This is associated with an increased risk of developing

ventricular arrhythmias and consequently, sudden cardiac death. Patient-specific 3D computational modelling and simulations can be used to predict non-invasively the

reentry circuits causing ventricular tachycardia (VT). Cellular automata (CA) elec- trophysiological models allow to reproduce VT while performing simulations near

real-time, overcoming the computational burden limitations of biophysical models. The aim of the present study was to create computational cardiac models capable of stratifying VT inducibility in infarcted patients by virtually applying the real pacing

protocol followed in the clinic using a novel CA-based solver developed at Universi- tat de València, in which no real clinical data has been tested yet. 3D computational

LV and biventricular models were obtained from cardiac magnetic resonance images provided by Centro Médico Teknon, allowing to identify the scar configuration and arrhythmogenic substrate of each patient. The models were reconstructed to fit in the CA, and seven pacing sites were defined to apply the virtual pacing protocol.

The obtained in-silico simulations results were compared with the actual results ob- tained by patients during an electrophysiological study (EPS). The similarity of the

results between in-silico and EPS demonstrated that the novel CA-based fast elec- trophysiological simulator together with the implementation of real pacing protocols

were valid for assessing VT risk in infarcted patients.