Attention-Based VGG-Residual-Inception Module for EEG-Based Depression Detection

Depression is a prevalent factor contributing to the increasing instances of suicide globally. Consequently, there is a pressing need for effective diagnosis and therapeutic interventions to alleviate depression symptoms. One potential tool for assessing depression levels is the electroencephalogram (EEG), a device that records and measures the brain's electrical activity. Previous studies have demonstrated the potential of using EEG data and deep learning models to diagnose mental disorders, paving the way for better comprehension and treatment of depression. As a result, this study offers a novel attention-based visual geometry group-residual-inception module (A-VGGRI) for classifying EEG data from healthy and major depression disorder people. The Patient Health Questionnaire-9 score is utilized to measure the depression level in this case. A-VGGRI's performance is examined using a depression dataset; the findings obtained by A-VGGRI have an accuracy of 96.35% and an area under the receiver operating characteristic curve of 0.96, demonstrating its usability in medical and industrial applications.