Analysis of the Effect of Cross-Frequency Coupling in the Diagnosis of Depression Using Resting State EEG Signal

Depression is currently a prevalent mental illness and is recognized as a social problem worldwide, characterized by low mood and impaired functioning. Therefore, the accurate and early identification of depression is one of the current challenges. The objective of this study was to investigate the effect of three types of Cross-Frequency Coupling (CFC) in the diagnosis of depression. Since any type of brain dysfunction can be manifested in Electroencephalogram (EEG) signals, in this study, EEG signals from 19 channels during a resting state were utilized, consisting of data from 22 depressed patients and 15 healthy individuals. The three types of CFC, including Phase-Amplitude Coupling (PAC), Phase-Phase Coupling (PPC), and Frequency-Amplitude Coupling (FAC), were computed within each electrode and between all electrode pairs. Ten features with the highest statistically significant differences were selected and fed into five machine learning classifiers: Support Vector Machine (SVM), Linear Discriminant Analysis (LDA), k-Nearest Neighbors (KNN), logistic regression (LR) and Decision Tree (DT). The KNN classifier, utilizing a combination of five features extracted from the FAC measure, achieved a classification accuracy of 100%. Therefore, this approach can be considered as an auxiliary tool for psychiatric purposes.