Localized cortical dipole imaging using a small number of electrodes based on independent component analysis.

The spatial resolution of scalp potential mapping is limited because of low conductivity of a skull. Cortical dipole layer imaging has been proposed as a method to visualize brain electrical activity with high spatial resolution. According to this method, about 100 electrodes were required to measure whole brain electrical activity. In the present study, we investigated simplified cortical dipole imaging with a small number of electrodes. The density of electrodes and the spatial resolution are in a trade-off relation. Thus, the number of electrodes was reduced by limiting the visualization region of interest, without lowering the density of electrodes. Moreover, independent component analysis was applied to the multiple signal sources to extract an attention signal from the other signals and noise. In simulation, even if the number of electrodes was reduced to 25, the obtained results were almost equivalent to the case with whole brain electrodes. The proposed method was applied to human experimental data of movement-related potential. We confirmed that the proposed method provided high resolution cortical dipole imaging with localized distribution.