Graph theoretical analysis of EEG after audiovisual stimulation in different anxiety states

Abstract

In this study, a brain network was created using graph theoretical analysis based on electroencephalography (EEG) data. The purpose of the study was to investigate the functional connectivity of the brain in different states of anxiety. Seventeen adults with anxiety (A-G), and 13 adults without anxiety (AF-G) were examined. They were given three different stimulations: resting, pleasant, and unpleasant. EEG was measured immediately after the stimulation. The EEG was analyzed by Fast Fourier Transform (FFT), coherence analysis, and graph theory. The results of FFT and coherence analysis showed that the anxiety group (A-G) had higher power spectra and coherence values than those for the anxiety-free group (AF-G) in all sessions. The results of graph theory analysis showed that the clustering coefficient and small-worldness in A-G were lower than those in AF-G, although the characteristic path length in A-G was higher than that in AF-G. This study shows that the brain of A-G has smaller clusters and longer paths to compare with those of AF-G. These events suggest that the brain of A-G would have an inefficient network structure to transmit emotional information.