Research On Brain Networks In Working Memory State Based On EEG

The human brain,with the sophisticated features in information processing,helps people become "the soul of the universe".As a nondestructive and noninvasive imaging technique,electroencephalogram(EEG)becomes a very important tool for people to understand the brain activity.In this thesis,the brain functional networks in resting state and working memory states under different loads are studied based on EEG.In this thesis,the experiments based on modified Sternberg working memory task is designed firstly,some subjects are chosen to finish the working memory tasks under different loads,and the EEG data are recorded synchronously.After the data preprocessing of the EEG data,the brain functional network is analyzed.First different brain regions are defined as the locations of different leads,the mutual correlation coefficients between each signal are calculated.Then the mutual correlation coefficient matrix is changed into a binary matrix by choosing appropriate threshold,brain function network connection diagram is set up based on binary matrix and the lead locations,and the distribution and dense degree of brain function network connection in different states are analyzed.Furthermore,the core brain regions responsible for different working memory tasks are found by calculating the correlation coefficient of brain network,and the dynamic changes of these regions under different states are analyzed.Finally,the EEG signals are decomposed into different frequencies using the wavelet packet decomposition method and the brain network changes in different memory states and different frequencies are analyzed.The results show that,the tightness of the brain network connection has significant differences in different task states,and presents a certain proportional relationship with the change of the load quantity.The frontal region and the occipital region are closely related to the completion of working memory tasks.These indicate that the human brain can make an accurate and rapid response to different experiment tasks,and it can activate different functional brain regions according to the complexity of tasks.