Electroencephalography (EEG) provides a direct measure of postsynaptic potentials with millisecond temporal resolution, and a means of studying the high temporal dynamics of functional networks. Approaches to estimating functional connectivity in EEG at the sensor level have been confounded by the diffusion of the EEG signal by the skull, however advances in source localization [17] have made it possible to minimize these confounds. Although the field is still in its infancy, several groups have begun to examine rsFC using measures of lagged connectivity between EEG source estimates. In applying this method, exact Low Resolution Electromagnetic Tomography [eLORETA; 17] – a linear inverse solution – is first used to compute the distribution of current density across voxels in the brain. Next, connectivity between intra-cortical sources is computed using lagged phase synchronization. This measure corrects for the effects of volume conduction as it represents the connectivity of two signals after the potentially artifactual zero-lag contribution has been excluded. Importantly, it can be applied to filtered data, allowing for the decomposition of connectivity at individual frequencies.
