Bioactive Functional Brain Network Research Based On Graph Theory Analysis

The brain is the central system for motor control,emotion and cognition.In the brain,the functional network and physical network of neurons,neural assembles,brain nucleis and brain areas are the physiological basis of brain functions.Hence,studies on brain functional network,to investigate organization modes of different state of brain functional network,are very meaningful and important for brain functions understanding,brain disease prevention/treatment and development of brain computers.In recent years,with the introduction of network science concepts to neuroscience,more and more neuroscientists started using the graph theory(mathematical tools of network sciences)to analyze the brain functional network.By graph theory analysis,researchers can use some quantitative values to study the functional connectivity between active units of the brain at the system level.Graph theory greatly promotes the progress of the study on brain functional networks.At present,most of the brain network research focused on the macro scale,the researches mainly used EEG,MEG and fMRI to assess brain activity at different sites.The graph theory analysis has been used in studying on intelligence of brain,the task performing brain and aging brain,to reveal the fundamental mechanism in network level.Besides,the graph theory would help to develop some tools to estimate the diseases of nervous system(e.g.Schizophrenia,Senile dementia and depression)and providing a reference for early screening and treatment of diseases.However,these macroscopic studies have some methodological limitations.First,they detect the direct or indirect synthetic reaction of neuronal populations.Electroencephalography(EEG)is an electrophysiological monitoring method to record synthetic electrical activity of the neurons under the scalp.Functional magnetic resonance imaging(fMRI)measures brain activity by detecting changes associated with blood flow.To study the brain functional network by these macroscopic technologies,it is need to be assumed that the information transfer could be represented by these technologies,but lacking the more direct evidences from neurons to be validated.Besides,these technologies have limitation on the spatio-temporal resolution and cost.In fact,functional brain network research can be more fire to reveal the functional organization of the brain in the mesoscopic scale(i.e.tens of and hundreds of neurons).It has been reported that,in the mesoscopic scale,the neural network in vitro,in the brain slices or in the monkey brain would demonstrate "small world"(most of the nodes were not connected directly but can communicate with few intermediate relay steps)or "rich-club"(nodes rich in connections tend to form strongly interconnected clubs)of functional network organization.However,brain functional network studies based on the graph theory in the mesoscopic scale are relatively preliminary.It is still unclear that how the external stimulation or different experimental design would influence the functional organization mode of the neural network.In this paper,we have studied the brain functional networks on three level,which are sμmmarized as follows: Part 1: Effects of drugs and electrical stimulation on functional connectivity of neural networks in vitro.Experiment 1: The effect of acute glutamate or GABA treatment on the functional connectivity and network topology of cortical cultures.We treated the neural cultures with glutamate of GABA,to investigate the influence of glutamate of GABA on the functional organization mode of functional connectivity in neural cultures.In this section,we acutely treated a mature cultured neural network on MEAs with accμmulating amounts of glutamate or GABA and recorded their altered electrophysiology.Subsequently,a cross-covariance analysis was applied to process the spiking activity in the network and to evaluate the connections between neurons.Finally,graph theory was used to assess the functional network property under acute glutamate treatment.Our results showed that glutamate increased the similarity,connectivity weight,density,largest component size and the small-worldness.Meanwhile,the similarity,network density,and largest component size were deceased by GABA.However,the connectivity weight and small-worldness were not found to be influenced by the acute GABA treatment.In this section,the brain functional network graph theory analysis system was built based on the muti-electrodes recording.Besides,it was suggested that,by using the graph theory,the influence of drug on neural cultures could be investigated by changes of functional organization mode.Experiment 2: Effect of chronic electrical stimulation on the functional connectivity of neural networks in vitro.We chronic electrical stimulating the neural cultures,to investigate the influence of electrical stimulation on the functional organization mode of functional connectivity in neural cultures.In this section,cortical cultures were electrically stimulated at a frequency of 0 Hz(control),0.02 Hz,and 0.2 Hz,between 7 and 26 days in vitro(DIV).The spontaneous activity of cortical cultures was recorded using MEAs.Next,a cross-covariance method and graph theory were applicable to investigate the dynamic changes in functional networks induced by electrical stimulation.Our results showed that over 3 weeks of stimulation,the network density(percentage of realized connections)significantly increased with maturation in the control and 0.02 Hz stimulation groups.However,the network density did not increase with culture age in the 0.2 Hz stimulation group.Moreover,all the cultures had a small-world topology at 14,18,22,and 26 DIV,and would not be affected by electrical stimulation.We found an asymmetry effect,in that partial electrical stimulation inhibited the formation of node connections in stimulated areas.This effect was more pronounced at 0.2 Hz than at 0.02 Hz stimulation.It was therefore suggested that the 0.02 and 0.2 chronic electrical stimulation would inhibit the connection formatting.Part 2: The influence of electromagnetic pulse stimulation on the functional connectivity of neural network in brain slicesWe electromagnetic pulse stimulate the cerebellar slice,to investigate influence of electromagnetic pulse stimulation on the functional organization mode of functional connectivity in the cerebellar slice.In this section,we applied 1Hz and 20 Hz repeated electromagnetic pulse stimulation on cerebellar slice,and recorded the neural network activity before and after stimulation by MEA.Next,a cross-covariance method and graph theory were applied to investigate the dynamic changes in functional networks induced by repeated electromagnetic pulse stimulation.Our results show that the 20 Hz electromagnetic pulse can significantly reduce the connectivity weight,average and network density of the functional network in cerebellar slice,but the small-worldness of the function network is not affected.1Hz of the electromagnetic pulse did not affect connectivity weight,average,network density and small-worldness of the functional network in cerebellar slice.It was concluded that the high frequency electromagnetic pulse stimulation would inhibit functional connectivity by reducing the connectivity weight,degree and network density.Part 3: The change of functional connectivity of functional network in cerebral cortex during the process of reach and grasp.To investigate the features of a functional network in cerebral cortex during the process of reach and grasp,monkeys were trained to perform the reach-grasp task.In this section,we record the electrophysiological activities of rhesus monkeys cerebral cortex by multi-electrode arrays during the process of reach and grasp.There were three multi-electrode arrays in S1,M1 and PPC brain regions respectively.Then we used the cross-covariance to evaluate the correlation of neural activity between any two channels.Finally,based on the threshold method,the functional networks were constructed,and the corresponding functional networks of hold,reach and grasp were analyzed.The results show that there is a specific relationship between the functional network topology structure properties and the motion type.Particularly,the connectivity weight of the hold motion was the lowest,the mean degree,the network density and the small world attribute were in the middle position;the reach motion had connectivity weight in the middle position,the lowest mean degree,the highest the network density and the lowest small-worldness.In grasp motion,its connectivity weight,mean degree and small-worldness were the highest,but its network density was the lowest.In addition,from nodes in different brain regions,we can get richer topology details,and there was a specific relationship between topology structure properties and movement too.It was suggested that we could found the movement features parameters in the brain functional network,such as connectivity weight,degree,network density and small-worldness.Above all,this study built a brain functional network graph theory analysis system based on mutielectrodes recording.And we performed the study on brain functional organization mode in culture neurons,brain slice and in vivo levels.It reveal the value of the brain functional network graph theory analysis to investigate the functional connectivity organization mode of neural network,under the different experimental design(e.g.drug,electrical stimulation,electromagnetic pulse stimulation,motor).It was also revealed that,performing study in a network review could provide a new way to understand the brain function.