A realistic 3-D FEM human torso model for electrocardiography applications

Summary form only given. During the past 30 years a lot of research effort has been directed toward the estimation of the heart electrical sources from the body surface potentials. This subject is now very well known as "inverse electrocardiography". The most widespread models for these sources can be identified as the "epicardial potentials" and the "equivalent dipoles or multipoles". Especially, the epicardial potentials seems to be very important for the clinicians, since they can identify the cardiac state (e.g. arrhythmias). The solution of the inverse problem is usually based on iterative solutions of the forward problem. Namely, for a given geometry, a conductivity distribution and a specified internal source, calculate the body surface potentials. Similarly, in the inverse problem the body surface potentials are measured, the conductivity and geometry are assumed known (physiological values) and the internal source is sought. This present work is restricted to the forward problem. What is in turn needed is a realistic 3-D volume conductor FEM model which accurately reflects the internal structure, but keeping in mind that this must be fast enough to be included within the inverse problem. For this purpose we have improved our previous model, with about 5-times finer mesh especially around the heart region. Some typical epicardial potential distributions and typical cardiac dipole sources are simulated and their effects on the surface potentials are presented. The next step is to include the present forward problem within an optimization scheme in order to solve the inverse one