Motor imagery EEG feature extraction method based on multi-feature fusion

doi:10.11772/j.issn.1001-9081.2019071167

Journal of Computer Applications ›› 2020, Vol. 40 ›› Issue (2): 616-620.DOI: 10.11772/j.issn.1001-9081.2019071167

• Frontier & interdisciplinary applications • Previous Articles

Motor imagery EEG feature extraction method based on multi-feature fusion

Fei LUO(), Pengfei LIU, Yuan LUO, Simeng ZHU

School of Optoelectronic Engineering，Chongqing University of Posts and Telecommunications，Chongqing 400065，China

Received:2019-07-03 Revised:2019-09-01 Accepted:2019-09-02 Online:2019-09-19 Published:2020-02-10
Contact: Fei LUO
About author:LIU Pengfei， born in 1992， M. S. candidate. His research interests include pattern recognition， emotion recognition.
LUO Yuan， born in 1972， Ph. D.， professor. Her research interests include intelligent signal processing， pattern recognition， artificial intelligence.
ZHU Simeng， born in 1999. His research interests include brain-computer interface， human-computer interaction.
Supported by:
the Chongqing Technology Innovation and Application Demonstration （Key Research in Industry） Project(cstc2018jszx-cyzdX0112)

多特征融合的运动想象脑电特征提取方法

罗飞(), 刘鹏飞, 罗元, 朱思蒙

重庆邮电大学光电工程学院，重庆 400065

通讯作者: 罗飞
作者简介:刘鹏飞（1992—），男，河南周口人，硕士研究生，主要研究方向：模式识别、情感识别
罗元（1972—），女，湖北宜昌人，教授，博士，主要研究方向：智能信号处理、模式识别、人工智能
朱思蒙（1999—），男，重庆人，主要研究方向：脑机接口、人机交互。
基金资助:
重庆市技术创新与应用示范（产业类重点研发）项目(cstc2018jszx-cyzdX0112)

Abstract

Abstract:

To solve the problems of low recognition rate and poor adaptability of single feature， a feature extraction method named Hilbert-CSP-Huang Transform （HCHT） was proposed based on Hilbert-Huang Transform （HHT） and Common Spatial Pattern （CSP）. Firstly， the Intrinsic Mode Function （IMF） was obtained by the Empirical Mode Decomposition （EMD） of original ElectroEncephaloGram （EEG） signals， and the IMF components were merged into a new signal matrix. Secondly， the time-frequency domain features were obtained by Hilbert spectrum analysis. Thirdly， the time-frequency domain features were extended into time-frequency-space features by further CSP decomposition of the constructed signal matrix. Finally， the feature set was classified by Support Vector Machine （SVM）. Experiments on the BCI Competition II dataset show that compared with methods based on HHT time-frequency and CSP spatial domain features， the proposed method has the recognition accuracy increased by 7.5， 10.3 and 9.2 percentage points respectively with smaller standard deviation. The online experimental results on the intelligent wheelchair platform show that HCHT can effectively improve the recognition accuracy and robustness.

Key words: ElectroEncephaloGram (EEG), Motor Imagery (MI), Hilbert-Huang Transform (HHT), Common Spatial Pattern (CSP), intelligent wheelchair

摘要：

针对单一特征识别率低、自适应性差等问题，提出一种基于希尔伯特-黄变换（HHT）和共同空间模式（CSP）的特征提取方法HCHT。首先，对原始脑电信号（EEG）进行经验模态分解（EMD）得到固有模态函数（IMF），并将IMF分量合并成新的信号矩阵；然后，对IMF进行希尔伯特谱分析，得到信号的时-频域特征；接着，对构造的信号矩阵进行进一步的CSP分解，将时-频域特征扩展成时-频-空域特征；最后，通过支持向量机（SVM）对特征集进行分类。在BCI Competition II数据集的实验表明，与HHT时-频域和CSP空域特征的方法相比，所提方法的识别准确率分别提高了7.5、10.3和9.2个百分点，且标准差更小。在智能轮椅平台进行在线实验的结果表明，HCHT能有效提高识别准确率和稳定性。

关键词: 脑电信号, 运动想象, 希尔伯特-黄变换, 共同空间模式, 智能轮椅

CLC Number:

TP242.6

Fei LUO, Pengfei LIU, Yuan LUO, Simeng ZHU. Motor imagery EEG feature extraction method based on multi-feature fusion[J]. Journal of Computer Applications, 2020, 40(2): 616-620.

罗飞, 刘鹏飞, 罗元, 朱思蒙. 多特征融合的运动想象脑电特征提取方法[J]. 《计算机应用》唯一官方网站, 2020, 40(2): 616-620.

Figures/Tables 10

Fig. 1 Single signal acquisition process

Fig. 2 Feature extraction process of HCHT

Fig. 3 EMD results of EEG indued by MI

Fig. 4 Comparison of single test accuracy

Tab. 1 Average accuracy

特征向量	平均准确率/%	标准差	特征向量	平均准确率/%	标准差
$F 1$	80.5	7.7	$F 3$	78.8	8.1
$F 2$	77.7	6.3	$F$	88.0	5.8

Tab. 1 Average accuracy

特征向量	平均准确率/%	标准差	特征向量	平均准确率/%	标准差
$F 1$	80.5	7.7	$F 3$	78.8	8.1
$F 2$	77.7	6.3	$F$	88.0	5.8

Tab. 2 Result comparison of multiple recognition methods

数据集	特征提取方法	分类方法	通道数	耗时/s	识别率/%
Data set III （Competition II）	香农熵+小波熵+样本熵^［11］	SVM	2	16.4	85.7
	EEMD+近似熵^［12］	线性判别分析	2	12.6	83.3
	本文方法HCHT	SVM	2	13.1	88.0
Data set I （Competition III）	CSSD+均值+Fisher判别分析^［17］	SVM	64	28.6	91.0
	自回归模型+功率谱+小波系数^［18］	正则化逻辑回归	33	25.8	87.0
	均值+功率谱^［19］	逻辑回归	不确定	22.4	86.0
	本文方法HCHT	SVM	2	12.9	87.3

Fig. 5 Electrode placement positions

Fig. 6 Experimental path

Fig. 7 Wheelchair trajectory curves based on three methods

Fig. 8 Time-consuming comparison of using HHT， CSP and HCHT to control wheelchair movement

References 19

1	WOLPAW J R， BIRBAUMER N. Textbook of neural repair and rehabilitation： brain-computer interfaces for communication and control［J］. Nature Reviews Neurology， 2017， 12（9）： 513-524. 10.1017/cbo9780511545061.036
2	王行愚，金晶，张宇，等. 脑控：基于脑—机接口的人机融合控制［J］. 自动化学报， 2013， 39（3）： 208-221. 10.3724/SP.J.1004.2013.00208
	WANG X Y， JIN J， ZHANG Y， et al. Brain control： human-computer integration control based on brain-computer interface approach［J］. Acta Automatica Sinica， 2013， 39（3）： 208-221. 10.3724/SP.J.1004.2013.00208
3	KIM M， KIM B H， JO S. Quantitative evaluation of a low-cost noninvasive hybrid interface based on EEG and eye movement［J］. IEEE Transactions on Neural Systems and Rehabilitation Engineering， 2015， 23（2）： 159-168. 10.1109/tnsre.2014.2365834
4	DJEMAL R， BAZYED A G， BELWAFI K， et al. Three-class EEG-based motor imagery classification using phase-space reconstruction technique［J］. Brain Sciences， 2016， 6（3）： No.36. 10.3390/brainsci6030036
5	BIJU K S， JIBUKUMAR M G. Ictal EEG classification based on state space modeling of intrinsic mode functions［J］. Procedia Computer Science， 2018， 125： 468-475. 10.1016/j.procs.2017.12.061
6	YÜCELBAŞ C， YÜCELBAŞ S， ÖZŞEN S， et al. Automatic detection of sleep spindles with the use of STFT， EMD and DWT methods［J］. Neural Computing and Applications， 2016， 29（8）： 17-33.
7	TARAN S， BAJAJ V. Motor imagery tasks-based EEG signals classification using tunable-Q wavelet transform［J］. Neural Computing and Applications， 2018， 31（11）： 6925-6932. 10.1007/s00521-018-3531-0
8	蔡淼，胡萍. 基于小波包变换的右手和脚运动想象任务分类［J］.中国医疗器械杂志， 2017， 413）：177-180. （CAI M， HU P. Task classification of right-hand and foot motion imagery based on wavelet packet transform［J］. Chinese Journal of Medical Instrumentation， 2017， 41（3）： 177-180. 10.3969/j.issn.1671-7104.2017.03.006
9	XYGONAKIS I， ATHANASIOU A， PANDRIA N， et al. Decoding motor imagery through common spatial pattern filters at the EEG source space［J］. Computational Intelligence and Neuroscience， 2018， 2018： No.7957408. 10.1155/2018/7957408
10	孙会文，伏云发，熊馨，等. 基于HHT运动想象脑电模式识别研究［J］.自动化学报， 2015， 41（9）：1686-1692. 10.16383/j.aas.2015.c150007
	SUN H W， FU Y F， XIONG X， et al. Identification of EEG induced by motor imagery based on Hilbert-Huang transform［J］. Acta Automatica Sinica， 2015， 41（9）： 1686-1692. 10.16383/j.aas.2015.c150007
11	CHEN S， LUO Z， GAN H. An entropy fusion method for feature extraction of EEG［J］. Neural Computing and Applications， 2018， 29（10）： 857-863. 10.1007/s00521-016-2594-z
12	杨默涵，陈万忠，李明阳. 基于总体经验模态分解的多类特征的运动想象脑电识别方法研究［J］. 自动化学报， 2017， 43（5）：743-752. 10.16383/j.aas.2017.c160175
	YANG M H， CHEN W Z， LI M Y. Multiple feature extraction based on ensemble empirical mode decomposition for motor imagery EEG recognition tasks［J］. Acta Automatica Sinica， 2017， 43（5）： 743-752. 10.16383/j.aas.2017.c160175
13	PFURTSCHELLER G， SCHLÖGL A. Data set III： BCI-experiment［EB/OL］. ［2019-04-08］. . 10.1088/1741-2560/2/4/l02
14	李明爱，崔燕，杨金福，等. 基于HHT和CSSD的多域融合自适应脑电特征提取方法［J］. 电子学报， 2013， 41（12）：2479-2486. 10.3969/j.issn.0372-2112.2013.12.025
	LI M A， CUI Y， YANG J F， et al. An adaptive multi-domain fusion feature extraction with method HHT and CSSD［J］. Acta Electronica Sinica， 2013， 41（12）： 2479-2486. 10.3969/j.issn.0372-2112.2013.12.025
15	张学军，黄婉露，黄丽亚，等. 公共空间模式算法结合经验模式分解的EEG特征提取［J］. 计算机工程与应用， 2017， 53（13）：9-15， 54. 10.3778/j.issn.1002-8331.1703-0199
	ZHANG X J， HUANG W L， HUANG L Y， et al. EEG signals feature extraction combined with empirical mode decomposition and common spatial pattern［J］. Computer Engineering and Applications， 2017， 53（13）： 9-15， 54. 10.3778/j.issn.1002-8331.1703-0199
16	QUITADAMO L R， CAVRINI F， SBERNINI L， et al. Support vector machines to detect physiological patterns for EEG and EMG-based human-computer interaction： a review［J］. Journal of Neural Engineering， 2017， 14（1）： No.011001. 10.1088/1741-2552/14/1/011001
17	魏庆国. 基于运动想象的脑-机接口分类算法的研究［D］. 北京：清华大学， 2006：11-16.
	WEI Q G. Studies on classification algorithms of motor imagery-based brain-computer interfaces［D］. Beijing： Tsinghua University， 2006： 11-16.
18	HAMMON P S， DE SA V R. Preprocessing and meta-classification for brain-computer interfaces［J］. IEEE Transactions on Biomedical Engineering， 2007， 54（3）： 518-525. 10.1109/tbme.2006.888833
19	BLANKERTZ B. BCI competition III — Final results［EB/OL］. ［2019-04-08］. . 10.1109/tnsre.2006.875642

[1]	HUO Shoujun, HAO Yan, SHI Huiyu, DONG Yanqing, CAO Rui. Pattern recognition of motor imagery EEG based on deep convolutional network [J]. Journal of Computer Applications, 2021, 41(4): 1042-1048.
[2]	CHANG Zheng, LUO Ping, YANG Bo, ZHANG Xiaoxiao. Respiratory sound recognition of chronic obstructive pulmonary disease patients based on HHT-MFCC and short-term energy [J]. Journal of Computer Applications, 2021, 41(2): 598-603.
[3]	HU Zhangfang, ZHANG Li, HUANG Lijia, LUO Yuan. Motor imagery electroencephalogram signal recognition method based on convolutional neural network in time-frequency domain [J]. Journal of Computer Applications, 2019, 39(8): 2480-2483.
[4]	HU Min, LI Chong, LU Rongrong, HUANG Hongcheng. Performance analysis of motor imagery training based on 3D visual guidance [J]. Journal of Computer Applications, 2018, 38(3): 836-841.
[5]	WANG Ruofan, LIU Jing, WANG Jiang, YU Haitao, CAO Yibin. Analysis algorithm of electroencephalogram signals for epilepsy diagnosis based on power spectral density and limited penetrable visibility graph [J]. Journal of Computer Applications, 2017, 37(1): 175-182.
[6]	HAN Min, SUN Zhuoran. Epileptic EEG signals classification based on wavelet transform and AdaBoost extreme learning machine [J]. Journal of Computer Applications, 2015, 35(9): 2701-2705.
[7]	Ke QI Dong-qing XIE. Steganalysis for JPEG MB2 steganography based on histogram differentiation random sequence [J]. Journal of Computer Applications, 2011, 31(04): 956-959.

Motor imagery EEG feature extraction method based on multi-feature fusion

多特征融合的运动想象脑电特征提取方法

RichHTML

PDF

Knowledge

Abstract

Cite this article

share this article

Figures/Tables 10

References 19

Related Articles 7

Recommended Articles

Metrics