EEG Classification Based on Multi-domain Features and Random Subspace Ensemble

doi:10.16182/j.issn1004731x.joss.19-0038

Abstract

Abstract: Aiming at the preprocessing feature extraction and classification recognition in BCI system, a method for EEG classification of motion imagery based on random subspaces ensemble learning of multi-domain features is proposed. Based on the analysis on the ERD/ERS characteristics of motion imagery (MI) signals, the multi-domain features of best effective time and frequency bands are extracted as the feature vectors, and the scale of the random subspace ensemble with cross-validation is adaptively chosen, and the EEG classification is realized by using linear discriminant analysis (LDA) classifiers ensemble. The test results show that the accuracy of the multi-domain features and random subspace ensemble can reach 90.71% and the Kappa coefficient can be 0.63,which are better than those of the first place in the competition, and thus prove the algorithm's effectiveness and progressiveness.

Key words: EEG signal, motion imagery, multi-domain features, ensemble learning, random subspace

CLC Number:

TP391.9

Deng Xin, Long Can, Mi Jianxun, Zhang Boxian, Sun Kaiwei, Wang Jin. EEG Classification Based on Multi-domain Features and Random Subspace Ensemble[J]. Journal of System Simulation, 2020, 32(9): 1787-1798.

References

[1] Wolpaw J R, Birbaumer N, Heetderks W J, et al.Brain-computer interface technology: a review of the first international meeting[J]. International Conference of the IEEE Engineering in Medicine and Biology Society (S1063-6528), 2000, 8(2): 164-173.
[2] 陈超, 平尧, 郝斌, 等. 基于脑机接口技术的写字系统建模仿真与实现[J]. 系统仿真学报, 2018, 30(12): 4499-4505.
Chen Chao, Ping Yao, Hao Bin, et al.Modeling, Simulation and Realization of Writing System Based on BCI Technology[J]. Journal of System Simulation, 2018, 30(12): 4499-4505.
[3] Chavez M, Grosselin F, Bussalb A, et al.Surrogate-Based Artifact Removal From Single-Channel EEG[J]. IEEE Transactions on Neural Systems and Rehabilitation Engineering (S1534-4320), 2018, 26(3): 540-550.
[4] 闫国强, 周宁宁, 张少白. 基于MMTD与小波硬阈值的脑电信号去噪方法[J]. 系统仿真学报, 2018, 30(4): 1490-1495.
Yan Guoqiang, Zhou Ningning, Zhang Shaobai.De-noising Method of EEG Signal Based on MMTD and Wavelet Hard-threshold[J]. Journal of System Simulation, 2018, 30(4): 1490-1495.
[5] Zhang Yu, Wang Bei, Jing Jin, et al.A Comparison Study on Multidomain EEG Features for Sleep Stage Classification[J]. Computational Intelligence and Neuroscience (S1687-5265), 2017: 1-8.
[6] Lin S, Guo S, Huang Z, et al.Determining AR order for BCI based on motor imagery[C]. Biomedical Engineering and Informatics. Shenyang: IEEE Press, 2015: 174-178.
[7] Song Y J, Sepulveda F.A Novel Technique for Selecting EMG-Contaminated EEG Channels in Self-Paced Brain-Computer Interface Task Onset[J]. IEEE Transactions on Neural Systems & Rehabilitation Engineering (S1534-4320), 2018(99): 1.
[8] 裴一飞, 杨淑娟. 运动想象脑电信号算法研究进展[J].北京生物医学工程, 2018, 37(2): 208-214.
Pei Yifei, Yang Shujuan.Research progress on motor imagery EEG signals[J]. Beijing Biomedical Engineering, 2018, 37(2): 208-214.
[9] Wang L, Xu G, Wang J, et al.Motor Imagery BCI Research Based on Sample Entropy and SVM[C]. International Conference on Electromagnetic Field Problems and Applications. Dalian: IEEE Press, 2012: 1-4.
[10] 孟建军, 盛鑫军, 姚林, 等. 基于共空域频谱模式的少通道运动想象分类[J]. 中国生物医学工程学报, 2013, 32(5): 553-561.
Meng Jianjun, Sheng Xinjun, Yao Lin, et al.Common Spatial Spectral Pattern for Motor Imagery Tasks in Small Channel Configuration[J]. Chinese Journal of Biomedical Engineering, 2013, 32(5): 553-561.
[11] Tan C Q, Sun F C, Zhang W C, et al.Spatial and spectral features fusion for EEG classification during motor imagery in BCI[C]. 2017 IEEE EMBS International Conference on Biomedical Health Informatics (BHI). Orlando: IEEE Press, 2017: 309-312.
[12] 王月茹, 李昕, 李红红, 等. 基于时-频-空间域的运动想象脑电信号特征提取方法研究[J]. 生物医学工程学杂志, 2014, 31(5): 955-961.
Wang Yueru, Li Xin, Li Honghong, et al.Feature Extraction of Motor Imagery Electroencephalography Based on Time-frequency-space Domanis[J]. Journal of Biomedical Engineering, 2014, 31(5): 955-961.
[13] 徐宝国, 宋爱国, 费树岷. 在线脑机接口中脑电信号的特征提取与分类方法[J]. 电子学报, 2011, 39(5): 1025-1030.
Xu Baoguo, Song Aiguo, Fei Shumi.Feature Extraction and Classification of EEG in Online Brain-Computer Interface[J]. Acta E1 Ecrromca Simca, 2011, 39(5): 1025-1030.
[14] Bhattacharyya S, Konar A, Tibarewala D N, et al.Performance analysis of ensemble methods for multi-class classification of motor imagery EEG signal[C]. International Conference on Control Instrumentation Energy Communication. Calcutta: IEEE Press, 2014: 712-716.
[15] 李明爱, 杨林豹, 杨金福. 具有在线自学习能力的脑电信号分类方法[J]. 计算机测量与控制, 2011, 19(11): 2763-2765, 2784.
Li Mingai, Yang Linbao, Yang Jinfu.A online Self--learning Approach to EEG Classification[J]. Computer Measurement & Control, 2011, 19(11): 2763-2765, 2784.
[16] Nicolasalonso L F, Corralejo R, Gomezpilar J, et al.Ensemble learning for classification of motor imagery tasks in multiclass brain computer interfaces[C]. Computer Science and Electronic Engineering Conference. Colchester: IEEE Press, 2014: 79-84.
[17] 王夏爽, 龚光红, 李妮. 视觉诱发脑电信号的处理研究[J]. 系统仿真学报, 2017, 29(增1): 146-154.
Wang Xiashuang, Gong Guanghong, Li Ni.Study and Processing of Visual Evoked EEG[J]. Journal of System Simulation, 2017, 29(S1): 146-154.
[18] Kevric J, Subasi A.Comparison of signal decomposition methods in classification of EEG signals for motor-imagery BCI system[J]. Biomedical Signal Processing and Control (S1746-8094), 2017, 31: 398-406.
[19] Jirayucharoensak S, Israsena P, Panngum S, et al.Online EEG artifact suppression for neurofeedback training systems[C]. Biomedical Engineering International Conference. Amphur Muang: IEEE Press, 2013: 1-5.
[20] Hosseini M, Hajisami A, Pompili D, et al.Real-Time Epileptic Seizure Detection from EEG Signals via Random Subspace Ensemble Learning[C]. International Conference on Autonomic Computing. Wurzburg: IEEE Press, 2016: 209-218.