A method for removal of eye movement artifacts from MEG - Simultaneous current source estimation of eyes and cortical activities from MEG and EOG data -

摘要

The magnetic fields generated by eye movements are major artifacts in MEG measurement. This report proposes a method for removal of eye movement artifact from MEG data by EOG data. In order to use EOG, one needs a EOG forward model that describes the transformation from eye current sources to EOG. We show that an EOG forward model and current sources can be simultaneously calculated by the hierarchical Bayesian estimation. A simulation experiment confirmed that the estimation of eye current sources can be improved. We further demonstrate that we can remove eye artifacts from real MEG data during smooth-pursuit eye movements.