Major Depressive Disorder Detection using Effective Connectivity of Resting-state EEG Signals and 2D-Convolutional Neural Network

Major Depressive Disorder (MDD) is a common mental health condition that can have a significant impact on an individual’s quality of life. This study is focused on the detection of MDD from resting-state Electroencephalogram (EEG) signals, using Effective Connectivity (EC) and 2D-Convolutional Neural Network (2D-CNN). MDD is a common mental health condition that is difficult to diagnose accurately through clinical assessment and self-reported symptoms. The clinical assessment and self-reported symptoms differ differently, and individuals may not always accurately report their experience due to various factors and thus, lead to inconsistencies in diagnosing MDD. The aim of this study is to explore the use of EC and 2D-CNN to overcome the challenges of utilizing complex characteristics of EEG signals for detection of MDD. A dataset of EEG recordings from 30 individuals with MDD and 30 HC, is used for this study. Overall, this study highlights the potential of EC and 2D-CNNs as a powerful tool for the analysis of complex EEG data and the detection of MDD. The results show that the proposed MDD diagnosis algorithm achieved 98.24% and 98.07% for sample-based and subject-based respectively in classifying MDD and HC test subjects.