Novel Biomarker Based on Hemispheric Differences in Brain Connectivity for Diagnosis of Depression

Biomarkers using EEG could be used to detect and analyze various mental diseases. Many researchers are working on biomarkers that are mostly based on spectral analysis, resulting in poor reproducibility so restricted to being used in a clinical environment. In this study, we proposed new biomarker, a potential biomarker for MDD using resting state EEG and EEG connectivity algorithm. To develop new biomarker, the $PRED+CT$ dataset consisting of 43 MDD patients and 73 healthy control was used. wPLI connectivity for the low alpha band was acquired from 60 EEG channels. An independent sample t-test was conducted, and only statistically significant connectivity ( $p < 0.001$ ) was selected among the connectivity. As a result, there was the biggest difference in connectivity between CPz and left hemisphere connectivity among MDD patients and healthy control. A decrease in connectivity between the central parietal region and frontal/temporal region located in the left hemisphere of MDD patients is also confirmed. New biomarker could be utilized in a diagnosis of MDD, and we expect it to be used in a clinical environment.