LAUSR

Abstract Functional activation as evidenced by blood oxygen level‐dependent (BOLD) functional MRI changes or event‐related EEG is known to closely follow patterns of stimulation or self‐paced action. Any lags are compatible with axonal conduction velocities and neural integration times. The important analysis of resting state networks is generally based on the assumption that these principles also hold for spontaneous fluctuations in brain activity. Previous observations using simultaneous EEG and fMRI indicate that slower processes, with delays in the seconds range, determine at least part of the relationship between spontaneous EEG and fMRI. To assess this relationship systematically, we used deconvolution analysis of EEG‐fMRI during the resting state, assessing the relationship between EEG frequency bands and fMRI BOLD across the whole brain while allowing for time lags of up to 10.5 s. Cluster analysis, identifying similar BOLD time courses in relation to EEG band power peaks, showed a clear segregation of functional subsystems of the brain. Our analysis shows that fMRI BOLD increases commonly precede EEG power increases by seconds. Most zero‐lag correlations, on the other hand, were negative. This indicates two main distinct neuromodulatory mechanisms: an “idling” mechanism of simultaneous electric and metabolic network anticorrelation and a “regulatory” mechanism in which metabolic network activity precedes increased EEG power by some seconds. This has to be taken into consideration in further studies which address the causal and functional relationship of metabolic and electric brain activity patterns. HighlightsBoth EEG and fMRI show characteristic fluctuations over time in the resting state. The functional significance of these fluctuations still unclear.By combining both methods and allowing time lags of up to 10.5 s between the two modalities, we show that functional MRI increases commonly precede EEG power increases by seconds.This indicates that at least one common mechanism connecting fMRI and EEG changes is much slower than expected with direct neural transmission.