Studies of localization of the ventricular preexcitation accessory pathway using heart model parameter optimization

The authors propose an approach to solve the electrocardiography (ECG) inverse problem in terms of heart model parameters. In this new method, a priori knowledge is used to limit the searching space of heart model parameters and the matching criteria are of vital importance. Here, the authors discuss the relations between the sites of ventricular preexcitation accessory pathways (AP) and the body surface potential maps (BSPMs) in detail based on the heart model. The results show that 3 features are of importance, i.e., (1) the minimum sites in BSPMs, (2) the correlation coefficients between BSPMs of the different AP sites, and (3) the areas (or lead number) of the negative and low-level positive potentials in BSPMs, are sensitive to the AP sites and relatively insensitive to different patients. Experimental results show that the resolutions of these characteristic parameters of BSPMs to recognize the AP site are 5-7 myocardial elements (MEs), 3-4 MEs and 1-3 MEs of the heart model, respectively. Using these characteristic parameters, 3 objective functions and the mathematical model of the optimization system were developed. The optimization algorithm is also described briefly. The experimental results indicate that this new approach to the ECG inverse problem is both feasible and efficient, and the localization accuracy can be obtained to be less than 3 myocardial elements (<4.5 mm) of the heart model.