SparNet: A Convolutional Neural Network for EEG Space-Frequency Feature Learning and Depression Discrimination

• DOI: 10.3389/fninf.2022.914823
• Publication Date: 2022-06-02T11:24:27Z
• Authors: Xin Deng; Xufeng Fan; Xiangwei Lv; Kaiwei Sun
• Keywords: Feature (linguistics); Depression; Convolution (computer science)

Depression affects many people around the world today and is considered a global problem. Electroencephalogram (EEG) measurement an appropriate way to understand underlying mechanisms of major depressive disorder (MDD) distinguish depression from normal control. With development deep learning methods, researchers have adopted models improve classification accuracy recognition. However, there are few studies on designing convolution filters for spatial frequency domain feature in different brain regions. In this study, SparNet, convolutional neural network composed five parallel SENet, proposed learn EEG space-frequency characteristics between The model trained tested by cross-validation method subject division. results show that SparNet achieves sensitivity 95.07%, specificity 93.66%, 94.37% classification. Therefore, our can conclude effective detecting using signals. It also indicates combination information identify patients with depression.