An estimation algorithm for neuromagnetic source distribution using MRI information

Abstract

We present an estimation algorithm for the location of magnetic sources in the human brain using magnetic resonance imaging (MRI) information. MRI information is used for limiting the estimation region to the cortex and analyzing acquired results. Magnetic fields of the brain are due to neural activity. As there are many neurons on the cortex, we regard it as a prime candidate for source location. The cortex is extracted from MR images. We assume that the points on the extracted cortex are current dipoles, and all the acquired points are considered as candidate sources. We then perform the estimation by means of the multiple signal classification (MUSIC) method. This algorithm can be considered as a first step in the estimation of the parameters of magnetic sources in the human brain. The acquired results may be interpreted as candidate locations at which sources exist, We perform computer simulations and apply this algorithm to real human brain data.