[1]潘赛虎,李文杰,等.基于共空间模式的运动想象脑电信号识别研究[J].南京师范大学学报(工程技术版),2014,14(02):055.
　Pan Saihu,Li Wenjie,Zhang Yi,et al.Classification of LeftRight Hand Motor Imagery Electroencephalogram Signals Based on a Feature Extraction Common Spatial Pattern Algorithm[J].Journal of Nanjing Normal University(Engineering and Technology),2014,14(02):055.
点击复制

基于共空间模式的运动想象脑电信号识别研究
分享到：

南京师范大学学报（工程技术版）[ISSN:1006-6977/CN:61-1281/TN]

卷:
14卷
期数:
2014年02期
页码:
055
栏目:
出版日期:
2014-06-30

文章信息/Info

Title:
Classification of LeftRight Hand Motor Imagery Electroencephalogram Signals Based on a Feature Extraction Common Spatial Pattern Algorithm
作者:
潘赛虎1; 李文杰1; 2; 张义1; 2
(1.常州大学信息科学与工程学院,江苏 常州 213164) (2.常州市生物医学信息技术重点实验室,江苏 常州 213164)
Author(s):
Pan Saihu1; Li Wenjie1; 2; Zhang Yi1; 2
(1.Faculty of Information Science and Engineering,Changzhou University,Changzhou 213164,China) (2.Changzhou Key Laboratory of Biomedical Information Technology,Changzhou 213164,China)
关键词:
运动想象; 脑-机接口; 特征提取; 模式识别
Keywords:
motor imagery; braincomputer interface; feature extraction; pattern classification
分类号:
R318.04;TP391.4
文献标志码:
A
摘要:
脑-机接口技术领域的关键问题是脑电信号的分类识别研究.本文针对脑电信号的分类问题,基于EGI-64导脑电采集系统得到7名被试者的左右手运动想象脑电数据,首先采用扩展InfomaxICA方法对脑电数据进行去噪处理;然后利用共空间模式方法对C3/C4 2个电极的脑电信号进行特征提取;最后比较了Fisher线性判别分析法、贝叶斯方法、径向神经网络和BP神经网络几种算法的平均分类率.结果表明:神经网络分类方法得到的平均分类率要高于其他2种方法,而BP神经网络方法的平均分类率最高,可以达到95.36%,但另外3种方法的运行速度明显高于BP神经网络.该结果为实时BCI系统实施提供了一定依据.
Abstract:
Classification of electroencephalogram(EEG)signal is an important issue in braincomputer interface(BCI).Based on the classification of the EEG signals,in this paper,we collect the leftright hand motor imagery EEG data of 7 subjects which are recorded by EGI-64 scalp electrodes placed according to the international 10/20 system.Firstly,the EEG data are denoised with extend InfomaxIndependent Component Analysis(ICA);Secondly,C3 and C4 electrodes features are extracted by using Common Spatial Pattern(CSP);Finally,the average classification rates of Fisher Linear Discriminant Analysis(FLDA),Bayesian,Radial Basis Function(RBF)neural network and BP neural network methods are compared.The classification results show that the average classification rate of neural network is higher than the other two methods,and that the average classification rate of BP neural network can be up to 95.36%,but the other three methods of running velocity is obviously faster than the BP neural network.The results provide a basis for realtime BCI system implementation.

参考文献/References:

[1]Yongwoong J,Chang S N,YongJoo K,et al.Eventrelated(De)synchronization(ERD/ERS)during motor imagery tasks:implications for braincomputer Interfaces[J].International Journal of Industrial Ergonomics,2011,41(5):428-436.
[2]Zou L,Wang X G,Shi G D,et al.EEG feature extraction and pattern classification based on motor imagery in braincomputer interface[J].International Journal of Software Science and Computational Intelligence,2011,3(3):43-46.
[3]Dennis J M,Jonathan R W.Sensorimotor rhythm based braincomputer interface(BCI):model order selection for autoregressive spectral analysis[J].Journal of Neural Engineering,2008,5(2):155-162.
[4]赵丽,郭旭宏.基于运动想象的脑电信号功率谱估计[J].电子测量技术,2012,35(6):81-87.
Zhao Li,Guo Xuhong.EEG power spectrum estimation based on motor imagery[J].Electronic Measurement Technology,2012,35(6):81-87.(in Chinese)
[5]ChiaFeng L,Shin T,Chih I H,et al.Reorganization of functional connectivity during the motor task using EEG time frequency cross mutual information analysis[J].Clinical Neurophysiology,2011,122(8):1 569-1 579.
[6]李庆华,山拜·达拉拜,孙全富.基于自适应小波阈值的心电信号去噪算法[J].计算机仿真,2013,30(10):368-432.
Li Qinghua,Senbai Dalabaev,Sun Quanfu.Brigemassart policy of application of ECG based on wavelet threshold denoising[J].Computer Simulation,2013,30(10):368-432.(in Chinese)
[7]Ling Z,Ying C Z,Laurence T T,et al.Single trial evoked potentials study by combining wavelet denoising and principal component analysis method[J].Journal of Clinical Neurophysiology,2010,27(1):17-24.
[8]Bian W,Fang Y,Jicai Z,et al.A frequency temporal spatial method for motor related electroencephalography pattern recognition by comprehensive feature optimization[J].Computers in Biology and Medicine,2012,42(4):353-363.
[9]Fraiwan L,Lweesy K,Khasawneh N.Classification of sleep stages using multi wavelet time frequency entropy and LDA[J].Methods of Information in Medicine,2010,49(3):230-237.
[10]胡明霞.表面肌电信号识别特征提取的仿真研究[J].计算机仿真,2013,30(5):340-350.
Hu Mingxia.Research on feature extraction and classification of electromyographic signals recognition[J].Computer Simulation,2013,30(5):340-350.(in Chinese)
[11]Wei Y H.Fuzzy hopfield neural network clustering for singletrial motor imagery EEG classification[J].Expert Systems with Application,2012,39(1):1 055-1 061.
[12]吴婷,颜国正,杨帮华,等.基于有监督学习的概率神经网络的脑电信号分类方法[J].上海交通大学学报:自然科学版,2008,42(5):803-806.
Wu Ting,Yan Guozheng,Yang Banghua,et al. Electroencephalography classification based on probabilistic neural network with supervised learning in brain computer interface[J].Journal of Shanghai Jiaotong University:Natural Science Edition,2008,42(5):803-806.(in Chinese)
[13]Zou L,Duan S L,Ma Z H,et al.Singletrial event related potentials extraction by using independent component analysis[C]//2th International Conference on Biomedical Engineering and Informatics.Tianjin,2009:1-5.
[14]Zou L,Xu S K,Ma Z H,et al.Automatic removal of artifacts from attention deficit hyperactivity disorder electroencephalograms based on independent component analysis[J].Cognitive Computation,2013,5(2):225-233.
[15]Muhammad T A,Wataru M,Christopher J J.Employing spatially constrained ICA and wavelet denoising,for automatic removal of artifacts from multichannel EEG data[J].Signal Processing,2012,92(2):401-416.

备注/Memo

备注/Memo:
收稿日期:2014-04-16.
基金项目:国家自然科学基金(61201096)、常州市科技项目(CE20135060、CM20123006、CJ20130026)、青蓝工程资助.
通讯联系人:潘赛虎,工程师,研究方向:信号处理.E-mail:pansaihu@126.com
更新日期/Last Update: 2014-06-30