Ali Khadem - School of Electrical and Computer Engineering University of Tehran Tehran, Iran
Gholam-Ali Hossein-Zadeh - School of Electrical and Computer Engineering University of Tehran, School of Cognitive Sciences Institute for Research in Fundamental Sciences (IPM) Tehran, Iran

EEG and MEG are popular modalities for functional/effective brain connectivity estimation; however they suffer from Volume Conduction (VC) artifact. VC artifact whichis an instantaneous linear mixing phenomenon may fake significant electrode couplings that are not due to true braininteractions. An ideal brain connectivity measure must be robustto VC artifact in the sense that it must never yield significant electrode couplings due to VC of independent sources. There areno criteria to compare the robustness of different brain functional/effective connectivity measures to VC artifact in realEEG/MEG datasets. In this paper, we propose a novel measure called Robustness Index (RI) using two surrogate data generationapproaches to fill this gap. RI is estimated over both simulated data and real EEG dataset for four functional connectivity measures: the absolute value of Pearson Correlation Coefficient(CC), Mutual Information (MI), magnitude squared Coherence (Coh) and the absolute value of Imaginary part of Coherency(ImC). RI on both datasets has correctly near %100 values for ImC which is theoretically robust to VC artifact. Also, for bothdatasets, the connectivity measures are ranked by RI as 1-ImC, 2-MI, 3-Coh and 4-CC which is consistent with their robustness levels to VC artifact.

Brain Connectivity Measures , EEG ، MEG , Robustness Index , Surrogate Data , Volume Conduction Artifact