Estimation and localization of electrical dipoles in somatosensory evoked potentials

Neuronal currents in the brain produce external magnetic fields and scalp surface potentials that can be measured using magnetoencephalography (MEG) and electroencephalography (EEG), respectively. In the context of the localization of neuronal sources, the forward problem is to determinate the potentials and magnetic fields that result from primary current sources. The inverse problem is to estimate the location of these primary current sources. Several algorithms have been applied to solve it. There are two kinds of such algorithms: Imaging methods, e.g. see R. D. Pascual-Marqui (1994), and parametric approaches. Here we present a revision of the MUSIC (J. C. Mosher et al., 1992) parametric algorithm and we analyze the effectiveness of this method with real EEG signals.