Evaluation of different cortical potential imaging methods using simulated EEG data

Different cortical potential imaging methods have been developed to directly link the scalp potentials with the cortical potentials. These methods make it possible to non-invasively investigate cortical activities with high spatial and time resolutions by using scalp EEG. However, although there are many different cortical potential imaging methods available, up to now, the accuracy and efficiency of these methods have not been rigorously evaluated nor compared. In this paper, we investigated a total of five different methods using ten different scenarios that employ simulated scalp EEG data with or without noise. Our results showed that 1), when only the center of electrical cortical activity needs to be estimated, single moving dipole and single dipole deviation scan methods are more accurate and more efficient than current density methods; and 2), with respect to current density methods, which are useful when the number of sources are unknown, the LORETA method with the L1-norm gives the highest accuracy, however, at a significant computational cost.