Depression Recognition with CNN Based on Feature Fusion and Attention Mechanism
It is of great practical significance to quickly and accurately identify,screen and early warn mild depression.By using EEG data and deep learning algorithm mental and psychological diseases can be machine-identified.A convolutional neural network(CNN)model based on feature fusion to effectively recognize depression.The attention mechanism is introduced into the CNN model to extract efficient spatio-temporal feature maps,enhance feature diversity and reduce the impact of individual differences.The results show that the average recognition accuracy of the model for depression reaches(99.39±0.14)%using EEG gamma rhythm.In addition,through the visual analysis of the convolutional layer feature map,the EEG differential electrodes of depression and normal subjects are ob-tained,and the depression classified with few electrodes,with the recognition accuracy of(91.41±1.11)%,showing that the deep learning model can effectively identify and screen mild depression.
国家科技期刊平台
NSTL
万方数据
