Attention EEG Research Based On Nonlinear Granger Causality

Attention is a neurocognitive process that describes a person’s mental and physical ability to focus on something.Attention deficit patients have symptoms of impulsiveness and hyperactivity,which makes it difficult to concentrate,which leads to a decline in learning and work efficiency,which seriously affects people’s normal life.Therefore,it is necessary to study the attentional EEG.In this paper,according to the attention EEG data of individuals with different attention levels in the three states of counting,eyes closed and idle,the non-linear Granger causality method of Sigmoid kernel,Gaussian kernel and Polynomial kernel is used to analyze the causal relationship between left brain and the right brain.The main research results are as follows:Firstly,Granger causality based on Sigmoid kernel for attentional EEG analysis.This paper selects the Granger causality based on the Sigmoid kernel function to analyze the causal relationship between the EEG signals of the left brain and the EEG signals of the right brain in three states: eyes closed,counting,and idle.This study obtained very good analysis results in the counting state.It is concluded that the directionality of the effect of the EEG signal of the left brain on the EEG signal of the right brain is greater than that of the EEG signal of the right brain on the EEG signal of the left brain under the counting state.Secondly,Granger causality based on Gaussian kernel for attentional EEG analysis.This paper selects the Granger causality based on the Gaussian kernel function to analyze the attention EEG signals in the three states,and verifies the conclusions obtained by using the Sigmoid kernel function in the counting state.The direction of the effect of the signal on the EEG signal of the right brain is smaller than that of the EEG signal of the right brain on the EEG signal of the left brain.In addition,it is also proved that individuals in the counting state have stronger causality between the left and right brains than in the eyes closed and idle states.Thirdly,Granger causality based on Polynomial kernel for attentional EEG analysis.This paper selects the Granger causality based on Polynomial kernel function to analyze the attention EEG signals in three states,and verifies the conclusions of the previous two kernel functions.The defects of Sigmoid kernel function and Gaussian kernel function in analyzing attentional EEG data are made up,and the causal relationship between left-brained attentional EEG and right-brained attentional EEG in three states is obtained.In addition,in order to verify the validity and reliability of the experimental conclusions,this paper uses SPSS statistical software to carry out independent samples T-test on the Granger causality index from left brain to right brain and right brain to left brain.Finally,for the convenience of storing the experimental results and adding,modifying and querying in the future,this paper is based on the Spring Boot and Spring Cloud frameworks,uses the java development language,and combines some advanced and mature front-end and back-end technologies and the popular Mysql database to implement a database management system.The system realizes the functions of login,data query and modification,new functions,online form design function,import and export of excel table function,user management,role management and other functions,which provide convenience for the input and storage of experimental data.