Aberrant Static and Dynamic Functional Brain Network in Depression Based on EEG Source Localization.

OBJECTIVE: Depression is accompanied by abnormalities in large-scale functional brain networks. This paper combined static and dynamic methods to analyze the abnormal topology and changes of functional connectivity network (FCN) of depression. METHODS: We collected resting-state EEG recordings from 27 depressed subjects and 28 normal subjects, then obtained 68 regions of interests (ROIs) by source localization. We took ROIs as the nodes and correlations as the edges to build FCNs and analyzed static network based on graph theory. We used a sliding window method followed by k-means clustering, states analyses and trend analysis of network metrics over time to study dynamic connectivity. RESULTS: The clustering coefficient (CC) and local efficiency in depression were increased, the characteristic path length and global efficiency were decreased, and local metrics had different manifestations in different resting state networks (RSNs); Depression had reduced connectivity in most RSNs, but increased connectivity in the default mode network, and there was a decoupling phenomenon between different RSNs; Depressed patients spent more time in sparsely connected states, their FCN's flexibility was less than normal subjects; The trend of CC over time was opposite between two groups. Most metrics in normal showed a relatively stronger correlation with time. SIGNIFICANCE: Our research may provide a deeper understanding of neurophysiological mechanisms of depression and new biomarkers for clinical diagnosis of depression.

EEG based depression recognition using improved graph convolutional neural network.

Depression is a global psychological disease that does serious harm to people. Traditional diagnostic method of the doctor-patient communication, is not objective and accurate enough. Thus, a more accurate and objective method for depression detection is urgently needed. Resting-state electroencephalography (EEG) can effectively reflect brain function, which have been used to study the difference of the brain between the depression patients and normal controls. In this work, the Resting-state EEG data of 27 depression patients and 28 normal controls was used in this study. We constructed the brain functional network using correlation, and extracted four linear features of EEG (activity, mobility complexity and power spectral density). We utilized a learnable weight matrix in the input layer of graph convolution neural network, creatively took the brain function network as the adjacency matrix input and the linear feature as the node feature input. We proposed our model Graph Input layer attention Convolutional Network (GICN), and it provided a good performance, showing the accuracy of 96.50% for recognition of depression and normal with 10-fold cross-validation, which indicated that our model could be used as an effective auxiliary tool for depression recognition. Besides, our method significantly outperformed other method. Additionally, the learnable weight matrix in the input layer was also used to find some edges and nodes that played an important role in depression recognition. Our findings showed that temporal lobe and parietal-occipital lobe had great effect in depression identification.