Brain Network Study Of Sleep Deprivation

Sleep deprivation(SD)refers to the sleep loss state caused by a variety of reasons,which would damage the individual's multiple emotion and cognitive abilities.However,the impact of SD on the mechanism of brain function still remains unclear.As a non-invasive vivo imaging technique,functional magnetic resonance imaging(fMRI)technology has been widely used in many areas of brain science research,which would also help deepen our understanding of the effects of SD on brain networks.Based on resting-state fMRI data,we proposed a group spatial sparse representation(group-SSR)model,and further used the dynamic functional connectivity(dFC)method to investigate the brain network of SD in this paper.In addition,based on the HCP dataset and simultaneous electroencephalography(EEG)-fMRI dataset,we further evaluated the impacts of global signal regression(GSR)on the dFC analysis,and investigated the potential neural source of global signal(GS).The main content of this paper is as follows:A method of sparse extraction of brain network of SD between groups.In this study,we presented a novel approach for making group inferences using sparse representation of resting-state fMRI data and its application to the identification of changes in functional networks in the brain following 36 hours of SD in contrast to the rested wakefulness(RW)stage.The results demonstrated that a number of functional networks with overlapping brain regions and multiple cognitive networks linked to SD were successfully extracted.Notably,the thalamus was observed to participate in multiple functional networks and exhibited distinct response patterns within these functional networks.These results provided some new insights into the impact of SD on functional organization of the brain by permitting spatial overlap between networks.Impacts of GSR on the analysis of dFC.In this study,a sliding-window based approach and a spectral clustering algorithm were used to evaluate the effects of GSR on characterizing functional connectivity(FC)dynamics as well as FC states.The results revealed that impact of GSR on dFC was spatially heterogeneous with some susceptible regions related to the vigilance network,and temporally modulated by the mean GS magnitude across windows.Furthermore,GSR seriously changed connectivity structures of the FC states with high spatial homogeneity.Finally,we reported the association between fluctuations in mean magnitude of the windowed GS and sliding-window measures of the anti-correlation between the default mode network and dorsal attention network as well as the time-varying EEG power within subjects.Overall,these observations suggest the potential neurophysiological sources in the GS dynamics that are likely linked to a spontaneous shift in vigilance or arousal states.We therefore highlight the careful use of GSR in fMRI analyses(e.g.,SD research)utilizing the sliding window correlation.Investigating the dynamics of brain network in SD.In this study,a sliding-window based approach and a spectral clustering algorithm were used to evaluate the effects of SD on the dFC of brain networks.Results revealed that SD resulted in significant changes in the time-averaging dFC across the whole brain as well as the temporal properties of FC states(e.g.,dwell time and transition probabilities).Moreover,the sleep state(RW vs.SD)can be effectively distinguished using the FC states' dwell time as well as the transition probabilities between states,demonstrating the potential links between these discrete states and individual behavior.