Frontiers in Human Neuroscience · 2018 · 24 citations · 64 references
Life SciencesNeurobiologyNeuroscienceCognitive NeuroscienceNeurophysiologySystems NeuroscienceBrain ImagingElectroencephalogram AnalysisElectroencephalographyFunctional ConnectivityFunctional NeuroimagingNeural Data AnalysisBlood Oxygenation Level-dependent ContrastBlood Oxygenation Level-dependent ImagingBrain Functional ImagingDefault Mode Network ConnectivityFunctional Brain ConnectivityFunctional Brain NetworksFunctional Magnetic Resonance ImagingFunctional Magnetic Resonance Imaging Correlates
Goal: We aimed to identify electroencephalographic (EEG) signal fluctuations within independent components (ICs) that correlate to spontaneous blood oxygenation level dependent (BOLD) activity in regions of the default mode network (DMN) during eyes-closed resting state. Methods: We analyzed simultaneously acquired EEG and fMRI eyes-closed resting state data in a convenience sample of 30 participants. Independent component analysis (ICA) was used to decompose the EEG time-series and common independent components (IC) were identified using data-driven IC clustering across subjects. The IC time courses were filtered into seven frequency bands, convolved with a haemodynamic response function and used to model spontaneous fMRI signal fluctuations across the brain. In parallel, group ICA analysis was used to decompose the fMRI signal into independent components from which the default mode network (DMN) was identified. Frequency and IC cluster associated haemodynamic correlation maps obtained from the regression analysis were spatially correlated with the DMN. To investigate the reliability of our findings, the analyses were repeated with data collected from the same subjects one year later. Results: Our results indicate a relationship between power fluctuations in the delta, theta, beta and gamma frequency range and the DMN in different EEG ICs in our sample as shown by small to moderate spatial correlations at the first measurement (0.234 < |r| < 0.346, p < 0.0001). Furthermore, activity within an EEG component commonly identified as eye movements correlates with BOLD activity within regions of the DMN. In addition, we demonstrate that correlations between EEG ICs and the BOLD signal during rest are in part stable across time. Discussion: We show that ICA source separated EEG signals can be used to investigate electrophysiological correlates of the DMN. The relationship between the eye movement component and the DMN points to a behavioral association between DMN activity and the level of eye movement or the presence of neuronal activity in this component. Previous findings of an association between frontal midline theta activity and the DMN were replicated.
64