Effect of excitation sequence of myocardial contraction on the mechanical response of the left ventricle

Abstract

In the past two decades there has been rapid development in the field of computational cardiac models. These have included either (i) mechanical models that assumed simultaneous myocardial activation, or (ii) electromechanical models that assumed time-varying myocardial activation. The influence of these modelling assumptions of myocardial activation on clinically relevant metrics, like myocardial strain, commonly used for validation of cardiac models has yet to be systematically examined, leading to uncertainty over their influence on the predictions of these models. This study examined the effects of simultaneous (mechanical), uniform endocardial, 3-patch endocardial (simulating the fascicles of the His-Purkinje system) and 1-patch endocardial (simulating the atrioventricular node) excitation sequences on the mechanical response of a synthetic human left ventricular model. The influence of the duration of the activation and time-to-peak contraction was also investigated. The electromechanical and mechanical models produced different strain distributions in early systole. However, these differences decayed as systole progressed. Using the same activation duration (74 ms) the average peak-systolic circumferential strain difference between the models was 0.65±0.37 %. A slightly prolonged activation duration (134 ms) resulted in no substantial difference increase (0.76±0.47 %). Differences up to 3.5 % were observed for prolonged activation durations (200 ms). Endocardial excitation produced non-physiological cumulative activation time distributions compared to the other models. Septal 1-patch excitation resulted in early systolic strain response that resembled pathological left bundle branch block. Decreased time-to-peak contraction exaggerated the effects of electrophysiology. The study found that excitation sequence minimally affects strain distributions at peak systole for physiological and even slightly pathological activation durations. However, electromechanical models with (patho)physiologically informed activation sequences are important for the accurate prediction of early systolic and pathological late systolic responses.