Study On Brain Function Network Of Depression Patients Based On EEG

Depression is a common mental disorder.The symptoms of depression include persistent emotional downturn,loss of interest in enjoyable activities,severe self-denial and suicidal tendencies.In recent years,as the quickening pace of life and the increasing pressure of work,more and more people are experiencing mild or moderate depression.In addition,compared to major depressive disorder,effective detection of mild depression at an early stage is still an urgent problem to be resolved that will have a significant sense.It has been possible for computer-aided diagnosis of mental illness possible with the development of computer technology.We can get more useful information based on data mining,which can provide a new perspective for the diagnosis of depression.At present,the results of the study on depression sometimes obtain inconsistent conclusions due to the choice of methods,the different experiment environment and the difference in mental state of the subjects,which will have an effect on diagnosis of depression using the findings as biological markers.The clinically important features of patients with depression are affective dysfunction and cognitive impairment.Both of these symptoms are associated with dysfunction in different brain regions.Therefore,our work based on brain functional network will be described as below:1)For the emotional facial expression pictures task,37 participants’ electroencephalogram(EEG)activity were recorded.Subjects were divided into normal group and mild depression group with 14 individuals in each group by BDI scale.Then coherence matrix was calculated between the 75 electrodes and the brain functional network was built by taking the threshold.Finally graph theory was used to study the "small-world" network properties of the two groups.The results of the study showed that the right brain region of the normal group cooperated more closely than the mild depression group in the emotional cognition process.And the activities of the prefrontal and parietal regions in the mild depression group were significantly lower than those in the normal group.At the same time,the functional networks of the mild depression group deviated from the “small-world” network properties in terms of feature path length.This study provides a new basis for the diagnosis and the validity of treatment for mild depression.2)For the resting state,EEG signal was recorded from 14 healthy subjects and 23 patients with major depression.The coherence was calculated and the functional network was constructed by MST to overcome the subjective bias caused by the selection of the threshold.Then the hierarchical clustering analysis of MST was conducted to study the differences in brain collaboration between two groups in the resting state.In order to ensure the reliability of the results,an EEG traceability study and re-coherence calculations was completed on the cerebral cortex.The EEG coherence of patients with MDD was significantly higher than that of the healthy controls in the θ band,especially in the left-hemisphere area.MST results showed that the depressive group had a higher ratio of leaf nodes.Compared with the normal group,patients with major depression lost a uniform cluster in the frontal area.Our findings indicate that the MDD group has stronger brain region interactions in the resting state and there is a left and right functional imbalance in the MDD group’s frontal area.This conclusion supports and complements the results of traditional brain network researches,and increases the reliability of research conclusions through horizontal comparison with traceability study.It provides strong support for using brain network properties as a basis for diagnosis and treatment of depression.