Effect Of Sleep Deprivation On Brain Network

Lack of sleep is the widespread social phenomenon, which affect our life and work. Through the sleep deprivation experiment, researcher found that human memory, emotions, logical thinking and other cognitive functions are affected. In addition, some cortex and subcutaneous tissue of brain changed significantly, including frontoparietal control area, temporal, hippocampus, amygdala, the central sulcus, inferior parietal angular gyrus and so on.Sleep deprivation is an experimental method of control subjects sleep time. By collecting fMRI data of participants in the sleep deprivation, using package of Statistical Parametric Mapping(SPM8), we preprocess the data by steps of head movement correction, spatial normalization, spatial smoothing. Then, using Anatomical Automatic Labeling(AAL) template, the brain is mapped to the 116 nodes of brain regions, we got the time series of each brain node. Next, we construct the functional brain network connectivity matrix by Pearson’s correlation coefficient. Finally, in the connectivity matrix, the value is range of-1 to 1, the positive connectivity stand for enhancement, in contrast, the negative connectivity is inhibition. Our study shows that the connectivity are mostly increased, and the Average connection, positive connection and negative connection is higher than the normal circumstances, which illustrates that integrally connectivity enhancement higher than weaken in sleep deprivation.According different threshold, we got the binary network of brain, and calculate the network properties of complex network. We find that clustering coefficient, characteristic path lengths and local efficiency is significantly increased, and the global efficiency is decreased. Characteristic path length enhancement and global efficiency decreased shows the collaboration reduce overall the brain regions, and the clustering coefficient and local efficiency are enhancements indicate that the local processing efficiency is increased. By contrast with the random network, the brain network is fit of small worldness between sleep deprivation group and normal sleep group, more importantly the small worldness is significantly increased in sleep deprivation group. Although the small worldness increased in sleep deprivation group, but the characteristic path length is increased and the global efficiency is decreased which indicates that our collaboration of brain is decline, this opposite situation suggesting that our brain has the compensation effect, our brain promote some brain region to prevent the influence of sleep deprivation.To further reveal the influence of the sleep deprivation, the two group difference in nodal efficiency was tested at the connection density of 0.41. We find that the nodal efficiency in anterior cingulate(ACG.R), inferior parietal(IPL.R), supramarginal gyrus(SMG.R), caudate nucleus(CAU.L, CAU.R), thalamus(THA.L) are significant decreased. And the temporal pole: middle temporal gyrus(TPOmid.L, TPOmid.R) are significant increased. We conclude that the nodal efficiency is affected by sleep deprivation obviously. For cerebellum, there are 11 brain regions significant change, and all changes is decreased, which indicates that our cerebellum is affected significantly. The cerebellum is in charge of balance-related functions, this may affect our daily life.Because of a lot of impact lack of sleep, it is easy to affect the safety of themselves and others, like fatigue driving. Many jobs need us stay awake, doctor, driver, astronauts and etc. In order to reduce the dangers of sleep deprivation, we use multi-voxel pattern analysis(MVPA) which can classify the sleep privation. Based on the obtained brain network connectivity matrix of two group. Taking the upper triangular matrix as classification features, and get a 4005-dimensional feature vector. Through Principal component analysis(PCA) and linear discriminant analysis(LDA) algorithm for dimension reduction, finally we find the PCA is the best method for MVPA. The MVPA get best generalization rate when select 185 brain features. By analyzing the 185 functional characteristics, found that these functional connectivity mainly in the prefrontal region of the brain, followed by occipital region. The temporal lobe, parietal region less functional connectivity distribution, indicating that the influence of sleep deprivation on the frontal region and the occipital region of the connection more strongly. The frontal area is mainly responsible for logical reasoning, language, spatial imaging and other advanced features, occipital has been primarily responsible for the human visual images and other advanced features.