Brain Functional Network Based on Mutual Information Analysis of EEGs and Its Application to schizophrenia

摘要

This paper presents a novel construction method of brain functional network by using spatial mutual information to estimate quantitatively correlation between different channels of electroencephalogram (EEG) signals. Firstly, 32-channel EEG signals were collected and the alpha rhythm of each signal was extracted by a wavelet packet filter, the spatial mutual information of different EEG channels was calculated respectively, the appropriate threshold was selected to construct the functional network. Then the research framework was applied to characterize the brain network difference of the schizophrenic patients and the control group. The result indicates that the network construction method proposed in this paper could portray different brain function states and it also can become a new effective method in analyzing and understanding the mechanism of patients with schizophrenia and other mental illnesses.