An End-to-End Deep Learning Model for EEG-Based Major Depressive Disorder Classification

Abstract

Major depressive disorder (MDD) is a prevalent mental illness associated with abnormalities in structural and functional brain connectivity, which has become a global public health problem. Early diagnosis is important and challenging for the treatment of MDD. Previous studies have proposed MDD classification methods based on brain connectivity features through functional connectivity and effective connectivity measures. However, artificial selection algorithms to compute brain connectivity features require prior knowledge and experience. Given that the representation learning capabilities of deep learning models and the ability to capture correlations between data of self-attention mechanism, this study proposed an end-to-end integrated DL model for classifying MDD patients and healthy controls based on the resting-state electroencephalography (EEG) data. The model first automatically learned the potential connectivity relationships among EEG channels through a multi-head self-attention mechanism, and then extracted higher-level features through a parallel two-branch convolution neural network module, and finally completed the classification through a fully connected layer. The leave-one-subject-out cross-validation method was utilized to evaluate the effectiveness of the proposed model on a publicly available EEG dataset. Ultimately, the average classification accuracy of the proposed model reached 91.06%, which was better than the comparison methods. The experimental results indicate that this study may provide a novel approach for brain connectivity modeling of MDD detection.

Keywords

• Major depression disorder
• electroencephalography
• convolution neural network
• brain connectivity
• self-attention mechanism