A Study on Depression Recognition Based on Cross-Frequency Coupling Brain Network Analysis

Depression is a common psychological condition that greatly impacts patients' quality of life. In this research, we introduced a brain network analysis method utilizing cross-frequency band coupling (CFC-BNA) for the identification of depression. We initially developed a brain network based on cross-frequency band coupling to identify alterations in the coupling of various frequency bands across the entire brain. The brain network was then binarised using the Prim minimum spanning tree algorithm and 12 network features were extracted for analysis and classification. The results showed an average accuracy of 77.36% for depression recognition using the Random Forest classifier. This approach emphasizes the intricate interactions of coupled neural circuits across different frequency bands in depression and holds promise for assessing depression.