Inter-subject variability of resting state brain activity explored using a data and model-driven approach in combination with EEG-FMRI

In this paper, we use co-registered EEG-fMRI during rest to investigate inter-subject-variability of BOLD signals in comparison with alpha-BOLD statistical parametric maps. A hierarchical clustering algorithm is used to detect spatial patterns of voxels showing correlated activity. The general- linear model is used to determine which of the identified patterns correlates significantly to the spontaneous variations of the alpha rhythm. For all sixteen subjects except one, the clustering of BOLD signal yielded very consistent regions wich included areas belonging to the "default mode" network and the neuronal networks involved in the generation of the alpha and mu rhythms. Furthermore, the BOLD clusters showed more consistency amongst subjects than the Alpha-BOLD statistical parametric maps obtained on a voxel-by-voxel basis. It is suggested that the larger inter-subject variability observed in the Alpha-BOLD statistical parametric maps when compared to the BOLD clusters is related to the individual variations in the EEG.