| As a fundamental phenomenon of life,sleep is important in our daily life associated with human health.Investigating the alterations of neural activity in brain networks during sleep can help us understand the underlying mechanism of the sleep-wake cycle.The default mode network(DMN)is the network related to the cognition and selfawareness,and was believed to play critical roles during sleep.However,traditional studies could not describe the relationships among multiple nodes since the functional connectivity of DMN is characterized by simple graphs.Recently,the hypergraph similarity network has been proposed to describe the higher-order interactions among nodes,which characterizes the whole functional structure of the brain network clearly without losing high-dimensional information.In this thesis,we constructed the hypergraph similarity network(hypergraph functional network)of electrophysiological DMN from rats.Then,we investigated the functional connectivity of DMN from the perspective of static network and dynamic network systematically,and revealed the changes of neural activity of DMN during the sleep cycle.The main results are as follows:(1)In different sleep states,the energy of each brain region of DMN was mainly concentrated in the low-frequency band.The power of SWS was found significantly increased in the Delta band and there were marked power changes in the Theta and Alpha bands during REM sleep.According to these findings,it can be concluded that oscillation patterns may be related to specific physiological functions of SWS and REM sleep.The results of the hypergraph functional network analysis indicated that opposite changes in functional connectivity between the hippocampus and other brain regions during SWS and REM sleep were at low frequencies.Previous studies suggested that the hippocampus was closely associated with memory processing in the brain.Thus,the formation of new memories during SWS may inhibit the activity of the hippocampus and parietal lobe during REM sleep.(2)Based on the research concerning the dynamic hypergraph functional network,the inherent functional modes in DMN were discovered.As sleep deepens,the number of functional modes decreased and no new modes appeared,suggesting that the complexity of the network has not increased.Furthermore,the anterior and posterior regions of DMN communicate in different ways during SWS and REM sleep could be observed.In the SWS state,the anterior and posterior DMN regions transmitted information indirectly through other modules,while they integrated directly in the REM sleep.Our results implied that the connectivity of the anterior and posterior DMN regions during SWS was significantly lower than that during REM sleep.(3)The conversion of DMN function mode analysis showed that the process of module reconstruction varied in different sleep stages.The original functional connectivity of DMN was shown broken and the small modules would be reassembled in the SWS state.Moreover,during REM sleep,the anterior and posterior DMN regions were initially integrated and the module with a higher integration degree would be gradually decomposed.Based on these results,we speculated the improvement of functional integration during REM sleep may be related to the recovery of consciousness.In conclusion,this study demonstrated that neural oscillations at low frequencies were an important way of information communication in the brain.Activities of the hippocampus and the temporal lobe played vital roles in module reconstruction and the information exchange between the anterior and posterior DMN regions.Besides,the intrinsic DMN functional modes reflect the multiple stability of the brain to some extent.Overall,our study demonstrated the changes of the hypergraph functional network of DMN in different sleep states and revealed the neural activity of DMN in sleep cycles.Our work may provide new insights for the research of sleep mechanism that possesses scientific significance as well as potential clinical value. |
PDF Full Text Request