Single-trial EEG classification of motor imagery using deep convolutional neural networks

详细信息    查看全文

• 作者：Zhichuan Tang^a ; ^b ; ^{ttzzcc@zju.edu.cn} ; Chao Li^b ; ^{superli@zju.edu.cn} ; Shouqian Sun^b ; ^{ssq@zju.edu.cn}
• 关键词：Motor imagery ; Non-invasive EEG ; Brain&ndash ; computer interface ; Deep convolutional neural network
• 刊名：Optik - International Journal for Light and Electron Optics
• 出版年：2017
• 出版时间：February 2017
• 年：2017
• 卷：130
• 期：Complete
• 页码：11-18
• 全文大小：1680 K
• 卷排序：130

文摘

Electroencephalogram (EEG) signal recorded during motor imagery (MI) has been widely applied in non-invasive brain–computer interface (BCI) as a communication approach. In this paper, we propose a new method based on the deep convolutional neural network (CNN) to perform feature extraction and classification for single-trial MI EEG. Firstly, based on the spatio-temporal characteristics of EEG, a 5-layer CNN model is built to classify MI tasks (left hand and right hand movement); then the CNN model is applied in the experimental data set collected from subjects, and compared with other three conventional classification methods (power + SVM, CSP + SVM and AR + SVM). The results demonstrate that CNN can further improve classification performance: the average accuracy using CNN (86.41 ± 0.77%) is 9.24%, 3.80% and 5.16% higher than those using power + SVM, CSP + SVM and AR + SVM, respectively. The present study shows that the proposed method is effective to classify MI, and provides a practical method by non-invasive EEG signal in BCI applications.