A method for two EEG sources localization by combining BP neural networks with nonlinear least square method

EEG source localization is well known as an important inverse problem of electrophysiology. Usually, there is no closed-form solution for this problem and it requires iterative techniques such as the Levenberg-Marquardt algorithm. However, the method requires long computing times, huge memory and large number of electrodes to avoid local minima. To overcome these problems, a method combining back propagation neural network (BPNN) with nonlinear least square method (NLS) is therefore proposed in this study. The new method shows how to estimate an approximate solution of the inverse problem by the BPNN method, and how to select the initial value of the NLS method due to the results of BPNN to obtain the optimum solution, where the problem is solved by POWELL iterative algorithm.