Automated Generation of Purkinje Networks in the Human Heart Considering the Anatomical Variability

Abstract

The Purkinje networks play a crucial role in the coordinated activation of the ventricular myocardium. The Purkinje networks grow in the right and left ventricles via the septum towards the apex and then upwards to the base with a different density. However, the inter-patient variability affects the growing pattern. In this study, we aimed to automatically create different Purkinje networks and evaluate the variability of their growth. We developed an algorithm that automatically generated ten different Purkinje network realizations in ten different ventricular geometries, obtaining 100 different Purkinje subendocardial networks. The growth was affected mainly by the size of the ventricles. Bigger ventricles (volume > 200 cm3) were highly populated with 1099 ± 43 Purkinje-Myocyte junctions compared to smaller ventricles (volume < 100 cm3) with 746 ± 38 Purkinje-Myocyte junctions. The Purkinje network activation sequence was also correctly verified by activating a biventricular geometry. In conclusion, we provide an algorithm that automatically produces biventricular Purkinje networks for any given ventricular geometry with a physiological activation.