Paper detail

Electrophysiological correlates of non-stationary BOLD functional connectivity fluctuations

Spontaneous fluctuations of the BOLD (Blood Oxygen Level-Dependent) signal, measured with fMRI (functional Magnetic Resonance Imaging), display a rich and neurobiologically relevant functional connectivity structure. This structure is usually revealed using time averaging methods, which prevent the detection of functional connectivity changes over time. In this work we studied the electrophysiological correlates of dynamical BOLD functional connectivity fluctuations, by means of long (approx. 50 min) joint electroencephalographic (EEG) and fMRI recordings. We identified widespread positive and negative correlations between EEG spectral power and fMRI BOLD connectivity fluctuations in a network of 90 cortical and subcortical regions. In particular, increased alpha (8-12 Hz) and beta (15-30 Hz) power were related to decreased functional connectivity, whereas gamma (30-60 Hz) power correlated positively with BOLD connectivity between specific brain regions. Furthermore, these patterns were altered for subjects undergoing vigilance changes, with an involvement of the slow delta (0.4 - 4 Hz) band in localized positive correlations. Finally, graph theoretical indices of network structure also exhibited sharp changes over time, with average path length correlating positively with alpha power extracted from central and frontal electrodes. Our results strongly suggest that non-stationary BOLD functional connectivity has a neurophysiological origin. Positive correlations of BOLD connectivity with gamma can be interpreted as increased average binding over relatively long periods of time, possibly due to spontaneous cognition occurring during rest. Negative correlations with alpha suggest functional inhibition of local and long-range connectivity, associated with an idling state of the brain.