Unbiased large-scale coherence mapping for simultaneously acquired EEG and fMRI data

The study of large scale interactions in the brain from EEG signals is carried out in the EEG community since years. However, the validity of a large scale parameter is limited by two factors [1]: the use of a non-neutral reference for the EEG recordings and the artifactual self-interaction between measured EEG signals introduced by volume conduction spread. In this paper, a novel approach for the study of large scale EEG coherency is proposed in which these biasing factors are eliminated. The artifactual self-interaction by volume conduction is eliminated by mapping interactions by means of the imaginary part of the complex coherency [2]; the bias introduced by the choice of an active reference site is eliminated by applying the Reference Electrode Standardization Technique (REST) [3,4] to scalp EEG recordings in order to approximately standardize the reference to a point at infinity that acts like a neutral virtual reference. The method is here applied to map coherency in the alpha band in the case of spontaneous activity EEG data acquired simultaneously to fMRI.