Dipole modelling of eye activity and its application to the removal of eye artefacts from the EEG and MEG.

The spatio-temporal dipole model approach has been used to identify the difference dipoles arising from changes in the ocular dipoles due to eye movements and blinks. Based on these results a method has been developed to remove eye artefacts from electrical or magnetic data. The method avoids distortions due to the head model by determining the spatial distribution of the signals from the eyes empirically. Using simultaneous modelling of the EEG or MEG activity with dipole sources distributed within the head together with the empirically determined spatial eye components, the eye activity can be estimated and removed from the EEG or MEG. This greatly reduces the distortion to the topography that is a concomitant of previous eye artefact correction methods. The advantages of the method are illustrated using simulated and real electrical data.