Functional Brain Networks for Sensory Maintenance in Top-Down Selective Attention to Audiovisual Inputs

Sensory maintenance in top-down selective attention to audiovisual inputs involves distributed cortical activations, while the connectivity between the widespread cortical regions has not been well understood. Graph theory has been demonstrated to be a useful tool in the analysis of brain networks. In this study, we used graph theoretical analysis to investigate the functional brain networks for sensory maintenance in top-down selective attention to audiovisual inputs. Electroencephalograms (EEGs) of 30 channels were recorded from 13 young healthy subjects during a passive view task and a top-down intersensory selective attention task. Phase synchronization indices of EEG signals in pair were computed to construct weighted brain networks. We found small-world properties of the brain networks during both passive view state and top-down selective attentional state in alpha, beta, and gamma bands. In addition, the significantly increased clustering coefficient and decreased characteristic path length were observed for brain networks during attentional state compared with passive view state in both beta band and gamma band. Our results suggest that functional brain networks in higher frequency bands, i.e., beta band and gamma band, are integrated in different ways during attentional state compared with passive view state.