Session S42.1

A Mesh-Less Method for Estimation of Left Ventricle Electrophysiological Activation Time

A Pashaei*, R Sebastian, V Zimmerman, BH Bijnens, AF Frangi

Universitat Pompeu Fabra
Barcelona, Spain

Modeling the dynamics of wave propagation in ventricular tissue is one of the challenging sides when studying electromechanical simulation of the heart. The number of variables in current physiological models, together with their particular mathematical requirements, makes the problem computationally expensive. Nonetheless, the level of detail required to model the dynamics of wave propagation in ventricular tissue highly depends on the ultimate target application, so that a rough estimation of the activation sequence is sometimes sufficient. This paper addresses the computational approach by using a mesh-less method for fast calculation of the activation time at an arbitrary point in the left ventricle. It is based on calculation of the time required to cover the path that the electrical signal would take to pass through. To account on the fast conduction system (higher conduction velocity), the myocardium is modeled as a multilayer domain. The path of the signal is estimated by finding the optimum trajectory of the wave passing a multilayer region. The model is described by two geometrical parameters and velocity map data. Geometric parameters are calculated from the surface mesh and the velocity map is approximated from experimental measurements obtained from the literature. The Purkinje network and fiber direction effects are accounted in the velocity map data. In order to check the performance of the method, the approach is compared with the results of Eikonal model in a two layer axisymmetric left ventricular domain. The comparison shows that this approach not only provides a fast estimation, but also obtains more accurate results using a two-layer left ventricular myocardium. This method is used for only one ventricle and although it does not contain the full detail electrophysiological interactions effects, it can model the main features of Purkinje network and fiber orientation in electrophysiological propagation. The mesh-less feature allows calculations for an arbitrary point, independently from the regional nodes, avoiding huge computations for unnecessary points. Phenomenological electrical cardiac models can help to roughly assess the effects of a given pacemaker for a large number of placements and settings in a fast way, and help to adjust a priori the settings in experiments that require complex models.

(Abstract Control Number: 113)