Noninvasive cardiac source imaging from potential and magnetic field data

Ventricular surface activation time maps are estimated from simulated and measured body surface potential (BSP) maps and extra‐corporal magnetic field maps. In a first step the transfer matrix, relating the primary cardiac sources to the measured potential and/or magnetic field data, is calculated applying the boundary element method. Activation times are determined by minimizing a cost function which is based on this transfer matrix. This optimization method is solved by a quasi Newton method. The critical point theorem is used in order to estimate the starting column matrix.