Brain Network Of Heroin-addicted Population Based On EEG

There are many types of drug addiction,the most concerned of which is heroin addiction.Heroin is currently one of the addictive drugs that threaten public health,so long-term use of heroin can cause very serious brain damage,which is usually irreversible.The study found that the biggest change in the brain caused by heroin is to increase the density of brain gray matter and the frequency of neuron activity is significantly reduced.These changes may cause addicts to have problems with decision-making and cognition.At present,many literatures focus on studying the addiction mechanism of addicted people from the perspective of event-related potential(ERP)and time-frequency analysis,but the use of brain network methods is still relatively small.Therefore,in order to further understand the risk decision-making disorder and impairment of cognitive function of heroin addicts,this article based on linear and non-linear methods,respectively made the following two aspects of research on the brain function network of heroin addicts:1.During the experiment of high and low risk decision-making,this paper used data from 34 subjects before the risk decision feedback result appeared 400 ms to 1000 ms after the decision feedback result appeared as the EEG data to be studied,in which the feedback result appeared before 400 ms is the reward expectation phase of the experiment,and 1000 ms after the feedback result appears is the experiment evaluation phase.In the experiment,17 heroin addiction group and 17 normal control group.Behavioral data results indicate that heroin addicts are more inclined to high-risk choices.Brain network results showed that during the expected reward period,the correlation and aggregation coefficients of the electrodes in the normal control group had obvious lateralization effects on the right hemisphere of the brain,while the lateralization effects disappeared in the heroin addiction group.At the same time,the study found that the heroin addiction group deviated from the "small world" attribute more than the normal group.In the result evaluation stage,the correlation between the two groups of people is related to the risk level and gains and losses.Combined with the minimum spanning tree(MST)analysis,in the reward expectation phase and theresult evaluation phase,the main brain areas that differ between the two groups of people are the right brain island,prefrontal lobe,and central apical area.There is a close relationship between the cognitive aspects of processing.The above results indicate that under two stages,the risk decision-making mechanism of heroin addicts has a certain degree of obstacles,so that when faced with risk choices,they cannot effectively evade risks and adjust the selection methods to avoid adverse risks.as a result of.2.Since the linear method can only analyze the time domain characteristics of EEG signals,it is in a different state:The EEG signals below have different characteristics.Therefore,in order to study the difference between the two groups of people more accurately,this article collected the EEG signals of 20 heroin addiction groups and 20 normal control groups in the resting state,and used nonlinear methods(mutual information method and phase synchronization method))Calculate the coherence between different brain regions.Build brain networks of the two groups of people on the four frequency bands Theta,Alpha,Beta,and Gamma,and calculate the corresponding network characteristics and analyze from the perspective of complex networks.The results showed that the global coherence of the normal control group was greater than that of the heroin addiction group under the four frequency bands.The statistical results found that there were significant differences in the three types of network characteristics of the two groups of people in the Alpha frequency band.The average aggregation coefficient and average node degree of the addiction group were lower than those of the normal control group,while the average path length was higher than that of the normal group.Then the AUC of the node efficiency,node centrality,and node degree of the two groups of people within a certain threshold range are calculated on the basis of the node degree,and take it as a local feature.The feature path lengths and aggregation coefficients of the two groups of people within the same threshold range are also calculated as global features.Finally,SVM is used as a classifier to obtain different classification accuracy under the combination of different features,and the optimal feature combination is obtained through comparison.On the one hand,this paper uses the method of brain network to study the mechanism of addiction and understand the brain dysfunction related to risk decision-making of heroin addicts,which provides new research for addiction Ideas.On the other hand,this paper uses the small-world attributes in the brain network asfeatures for classification to obtain the best combination of classification features.This provides a new direction for the further diagnosis and withdrawal of heroin addiction.