| Sleep deprivation (SD) is a general lack of the necessary amount of sleep. It occurs when an individual does not sleep during the normal sleep-wake cycles. SD has a broad variety of effects on human performance and neural functions. Research on SD effects is a powerful method to investigate the mechanisms underlying sleep regulation and to provide references for countermeasures. In the last two decades, SD has been studied by psychologists and military research institutes. Electroencephalography (EEG) can record synchronized firing behavior of neural assemblies in milliseconds so that EEG analysis has become a powerful tool for measuring brain states during cognition. In previous studies, analysis of EEG data was only used to reveal frequency-specific or time-specific changes under SD state. There was lack of integrated research for all domains, and in addition, communications between functional areas of the brain need to be studied and the working mechanism of SD needs to be further revealed.By the study of EEG features induced by SD, this paper tried to evaluate the effects of SD and to further reveal its working mechanism. There were four parts of work, as following,Firstly, we designed experiment paradigms and finished tests under two states including normal state and one night's sleep deprived state. Basic EEG features in time-domain and frequency-domain were studied. Results showed that after SD, the amplitude of P300 component decreased while the latency increased in auditory and visual event-related potentials, revealed a reduction of brain response ability. The increase of theta activities and decrease of alpha activities of spontaneous power spectrum showed a reduction of brain activity.Secondly, EEG data were decomposed in temporal, frequency and spatial domains to evaluate the brain state. The analysis of EEG data by wavelet transformation, PARAFAC decomposition and statistical analysis were used to identify where the differences appeared in three domains. Results showed that in auditory process the changes of brain states caused by SD appeared around 40Hz, about 400ms after stimulation on prefrontal and frontal lobes. Negative effects on neuronal activity and oscillation were observed.Thirdly, phase coherence analysis was applied on wavelet transformed EEG, to evaluate the effects of SD on synchronization of brain oscillation and to help understanding the working mechanism of SD. Results showed that after SD, the activities of inter-trial phase coherence decreased and delayed on prefrontal lobe, indicating the reduction of the ability of brain for maintaining its state to respond to continuous stimuli; activities of evoked response phase coherence between prefrontal and occipital lobes subsided, indicating the information transfer between functional cortex areas lessened.The last one, approximate entropy (ApEn) was calculated to assess the non-linear influence on brain functional states, and to exploit non-linear changing rules. ApEn of EEG data were calculated and brain information map (BIM) was reconstructed. Results showed that after SD, ApEn of spontaneous EEG minified over the whole brain by various degrees; while the asymmetry of prefrontal lobes changed by the center of high complexity shifted from left to right hemisphere, indicating prohibition of the brain.By BIM of ERP, ApEn on prefrontal cortex decreased while that on parietal lobes and temporal lobes increased, showing the compensation effect of cortex.To sum up, the analysis of basic EEG features verified the availability of the experiments, and provides a foundation for further analysis. By the study of time-frequency-space characteristics, phase coherence features and non-linear features of EEG data under two states, a relatively complete way to evaluate the effects of SD on brain cognition was formed. It may provide objective evidence to support previous conclusions, and lead to a better understanding of the working mechanism of SD.
|
PDF Full Text Request