基于噪声辅助快速多维经验模式分解的运动想象脑电信号分类方法

doi:10.11959/j.issn.2096-6652.202026

智能科学与技术学报 ›› 2020, Vol. 2 ›› Issue (3): 240-250.doi: 10.11959/j.issn.2096-6652.202026

基于噪声辅助快速多维经验模式分解的运动想象脑电信号分类方法

郑潜¹,乔丹¹,郎恂²,谢磊¹(),李东流³,王琪冰³,苏宏业¹

¹ 浙江大学智能系统与控制研究所，浙江杭州 310027
² 云南大学信息学院电子工程系，云南昆明 650091
³ 森赫电梯股份有限公司，浙江湖州 313000

修回日期:2020-08-20 出版日期:2020-09-20 发布日期:2020-10-20
作者简介:郑潜（1995- ），男，浙江大学智能系统与控制研究所博士生，主要研究方向为生物医疗信号处理、时频分析、机器学习等|乔丹（1995- ），女，浙江大学智能系统与控制研究所硕士生，主要研究方向为生物医疗信号处理、时频分析、因果分析等|郎恂（1994- ），男，云南大学信息学院电子工程系博士后，主要研究方向为工业过程性能评估、信号处理、时频分析等|谢磊（1979- ），男，浙江大学智能系统与控制研究所教授、博士生导师，主要研究方向为先进控制理论与应用、工业数据挖掘与人工智能、工业控制系统性能评估等|李东流（1966- ），男，高级工程师，森赫电梯股份有限公司董事长，浙江省电梯行业协会会长|王琪冰（1978- ），男，博士，正高级工程师，IET Fellow，森赫电梯股份有限公司研究院院长，英国英中智能机电装备与机器人联合研究中心首席科学家|苏宏业（1969- ），男，浙江大学智能系统与控制研究所教授、博士生导师，主要研究方向为工业智能与优化控制、机器人与智能无人系统等
基金资助:
国家重点基础研究发展计划基金资助项目(2018YFB1701102);中央高校基本科研基金资助项目

Classification of motor imagery signals using noise-assisted fast multivariate empirical mode decomposition

Qian ZHENG¹,Dan QIAO¹,Xun LANG²,Lei XIE¹(),Dongliu Li³,Qibing Wang³,Hongye SU¹

¹ State Key Laboratory of Industrial Control Technology,Zhejiang University,Hangzhou 310027,China
² Department of Electronic Engineering,Information School,Yunnan University,Kunming 650091,China
³ Sicher Elevator Co.,Ltd.,Huzhou 313000,China

Revised:2020-08-20 Online:2020-09-20 Published:2020-10-20
Supported by:
The National Key Basic Research and Development Program of China(2018YFB1701102);The Fundamental Research Funds for the Central Universities

摘要/Abstract

摘要：

脑机接口是一项新兴的技术，它可以处理分析采集到的运动想象脑电信号，从而实现对外部辅助设备的控制。针对目前运动想象脑电信号处理方法计算效率低、分类准确率不高等问题，提出了一种新的基于噪声辅助快速多维经验模式分解（NA-FMEMD）的运动想象脑电信号分类方法。该方法首先利用 NA-FMEMD 得到全部的多维本征模式函数和趋势项；接着，根据平均频率选取特定的信号层，构建出新的多维信号；然后，通过共空间模式提取出脑电信号的特征向量；最后，将特征向量输入支持向量机分类器中进行分类。分别采用仿真数据和BCI Competition IV数据进行测试，并与基于噪声辅助多维经验模式分解（NA-MEMD）的方法进行比较，验证了所提方法的有效性和优势。

关键词: 脑电信号, 运动想象, 噪声辅助快速多维经验模式分解, 共空间模式

Abstract:

The brain-computer interface is an emerging technology,which can analyze the collected motor imagery signals to control the external auxiliary equipment.A new method based on the noise-assisted fast multivariate empirical mode decomposition (NA-FMEMD) algorithm was proposed for electroencephalogram signal feature extraction and classification.The method outperformed state-of-the-art methods based on noise-assisted multivariate empirical mode decomposition in not only computational efficiency but also classification accuracy.Firstly,all multivariate intrinsic mode functions and trend signals were obtained by the NA-FMEMD.Secondly,the multivariate signals with specific frequency bands were selected by computing their average frequencies.Thirdly,the common spatial pattern was applied to extract features.Finally,the feature vectors were classified using a support vector machine.Simulation data and BCI Competition IV data are used to verify the effectiveness and advantage of the new method,and the method is compared with noise-assisted multivariate empirical mode decomposition.

Key words: electroencephalogram signals, motor imagery, noise-assisted fast multivariate empirical mode decomposition, common spatial pattern

中图分类号:

TP391

郑潜, 乔丹, 郎恂, 等. 基于噪声辅助快速多维经验模式分解的运动想象脑电信号分类方法[J]. 智能科学与技术学报, 2020, 2(3): 240-250.

Qian ZHENG, Dan QIAO, Xun LANG, et al. Classification of motor imagery signals using noise-assisted fast multivariate empirical mode decomposition[J]. Chinese Journal of Intelligent Science and Technology, 2020, 2(3): 240-250.

图/表 13

图1

图2

图3

图4

表1

表2

表3

图5

表4

图6

表5

表6

图7

参考文献 29

[1]	丁晓慧 . 运动想象脑电信号特征提取与分类研究[D]. 杭州:杭州电子科技大学, 2016.
	DING X H . Feature extraction and classification of motor imagery EEG signals[D]. Hangzhou:Hangzhou Normal University, 2016.
[2]	AKRAMI A , SOLHJOO S , MOTIE-NASRABADI A . EEG-based mental task classification:linear and nonlinear classification of movement imagery[C]// The 27th Annual International Conference of the Engineering in Medicine and Biology Society. Piscataway:IEEE Press, 2006: 4626-4629.
[3]	PARK C , LOONEY D , UR REHMAN N ,et al. Classification of motor imagery BCI using multivariate empirical mode decomposition[J]. IEEE Transactions on Neural Systems and Rehabilitation Engineering, 2013,21(1): 10-22.
[4]	SILVA G P H L D . Event-related EEG/MEG synchronization and desynchronization:basic principles[J]. Clinical Neurophysiology, 1999,110(11): 1842-1857.
[5]	NIKOULINE V V , LINKENKAER-HANSEN K , WIKSTR?M H ,et al. Dynamics of mu-rhythm suppression caused by median nervestimulation:a magneto encephalographic study in human subjects[J]. Neuroscience Letters, 2000,294(3): 163-166.
[6]	LOTTE F , CONGEDO M , LéCUYER A ,et al. A review of classification algorithms for EEG-based brain–computer interfaces[J]. Journal of Neural Engineering, 2007,4(2):R1.
[7]	RODRíGUEZ-BERMúDEZ G , GARCíA-LAENCINA P J . Automatic and adaptive classification of electroencephalographic signals for brain computer interfaces[J]. Journal of Medical Systems, 2012,36(1): 51-63.
[8]	PARK S A , HWANG H J , LIM J H ,et al. Evaluation of feature extraction methods for EEG-based brain–computer interfaces in terms of robustness to slight changes in electrode locations[J]. Medical ＆Biological Engineering ＆ Computing, 2013,51(5): 571-579.
[9]	ZHANG Y , LIU B , JI X M ,et al. Classification of EEG signals based on autoregressive model and wavelet packet decomposition[J]. Neural Processing Letters, 2017,45(2): 365-378.
[10]	FAUST O , ACHARYA U R , ADELI H ,et al. Wavelet-based EEG processing for computer-aided seizure detection and epilepsy diagnosis[J]. International Journal of Engineering Research and Applications, 2015,26: 56-64.
[11]	KRISHNA D H , PASHA I A , SAVITHRI T S . Classification of EEG motor imagery multi class signals based on cross correlation[J]. Procedia Computer Science, 2016,85: 490-495.
[12]	MA Y L , DING X H , SHE Q S ,et al. Classification of motor imagery EEG signals with support vector machines and particle swarm optimization[J]. Computational and Mathematical Methods in Medicine, 2016,2016: 1-8.
[13]	PFURTSCHELLER G , NEUPER C . Motor imagery and direct brain-computer communication[J]. Proceedings of the IEEE, 2001,89(7): 1123-1134.
[14]	HUANG N E , SHEN Z , LONG S R ,et al. The empirical mode decomposition and the Hilbert spectrum for nonlinear and non-stationary time series analysis[J]. Philosophical Transactions Mathematical Physical ＆ Engineering Sciences, 1998,454(1971): 903-995.
[15]	WU Z H , HUANG N E . Ensemble empirical mode decomposition:a noise-assisted data analysis method[J]. Advances in Adaptive Data Analysis, 2009,1(1): 1-41.
[16]	REHMAN N , MANDIC D P . Multivariate empirical mode decomposition[J]. Proceedings of the Royal Society A:Mathematical,Physical and Engineering Sciences, 2009,466(2117): 1291-1302.
[17]	UR REHMAN N , MANDIC D P . Filter bank property of multivariate empirical mode decomposition[J]. IEEE Transactions on Signal Processing, 2011,59(5): 2421-2426.
[18]	LANG X , ZHENG Q , XIE L ,et al. Direct multivariate intrinsic time-scale decomposition for oscillation monitoring[J]. IEEE Transactions on Control Systems Technology, 2019
[19]	LANG X , ZHANG Z , XIE L ,et al. Time-frequency analysis of plant-wide oscillations using multivariate intrinsic time-scale decomposition[J]. Industrial ＆ Engineering Chemistry Research, 2018,57(3): 954-966.
[20]	LANG X , ZHENG Q , ZHANG Z M ,et al. Fast multivariate empirical mode decomposition[J]. IEEE Access, 2018,6: 65521-65538.
[21]	WANG Z S , MAIER A , LOGOTHETIS N K ,et al. Single-trial classification of bistable perception by integrating empirical mode decomposition,clustering,and support vector machine[J]. EURASIP Journal on Advances in Signal Processing, 2008,2008: 1-8.
[22]	CHEN X , LIU A P , CHIANG J ,et al. Removing muscle artifacts from EEG data:multichannel or single-channel techniques[J]. IEEE Sensors Journal, 2016,16(7): 1986-1997.
[23]	BLANKERTZ B , DORNHEGE G , KRAULEDAT M ,et al. The noninvasive berlin brain-computer interface:fast acquisition of effective performance in untrained subjects[J]. NeuroImage, 2007,37(2): 539-550.
[24]	WANG J J , FANG XUE , LI H . Simultaneous channel and feature selection of fused EEG features based on sparse group lasso[J]. Biomed Research International, 2015
[25]	YUAN H , DOUD A , GURURAJAN A ,et al. Cortical imaging of event-related (de) synchronization during online control of braincomputer interface using minimum-norm estimates in frequency domain[J]. IEEE Transactions on Neural Systems and Rehabilitation Engineering, 2008,16(5): 425-431.
[26]	VAPNIK V N . Statistical learning theory[M]. New York: WileyPress, 1998.
[27]	BREIMAN L , FRIEDMAN J H , OLSHEN R A ,et al. Classification and regression trees[M]. Belmont: Wadsworth International GroupPress, 1984.
[28]	COVER T , HART P . Nearest neighbor pattern classification[J]. IEEE Transactions on Information Theory, 2003,13(1): 21-27.
[29]	DUDAR O , HARTP E , STORKD G . Pattern classification,2nd ed[M]. New York: WileyPress, 2001.

信号	β波/Hz	α波/Hz	θ波/Hz
x₁(t)	24.9/22.8/26.7/22.1	11.1/12.7/10.9/10.7	7.3/6.4/6.2/4.2
x ₂(t)	28.1/28.9/20.0/25.6	12.1/12.0/12.0/12.3	7.4/5.9/4.0/5.2
x ₃(t)	18.8/19.6/27.0/25.9	9.1/8.1/11.1/10.4	5.3/6.5/5.1/4.7
x ₄(t)	19.1/15.4/13.1/23.4	11.7/9.3/8.8/8.1	4.5/4.9/6.9/4.8

信号	d₁/Hz	d₂/Hz	d₃/Hz	d₄/Hz	d₅/Hz
x₁(t)	37.3/37.0/33.7/34.9	24.7/22.8/26.6/21.9	11.2/12.9/11.3/11.0	7.3/6.5/6.5/5.1	5.7/5.8/4.5/4.2
x ₂(t)	34.6/32.2/37.3/33.9	27.9/28.5/20.1/25.5	12.2/12.1/12.3/12.6	7.4/6.1/6.3/5.3	5.7/5.9/4.2/5.1
x ₃(t)	38.2/39.2/33.8/32.5	19.0/19.7/26.9/25.7	11.1/16.2/11.2/10.4	9.1/5.6/5.4/4.9	5.2/6.0/4.9/4.4
x ₄(t)	37.1/37.0/38.5/33.5	19.1/16.8/25.5/23.5	11.9/14.4/10.7/8.5	10.0/9.2/7.0/5.6	4.6/4.9/4.5/4.5

信号	d₁/Hz	d₂/Hz	d₃/Hz
x₁(t)	24.5/22.8/26.7/21.9	11.1/11.5/11.0/10.9	7.3/6.3/6.1/4.2
x ₂(t)	28.0/29.1/20.2/25.6	12.2/12.1/12.0/12.6	7.3/5.9/4.3/5.3
x ₃(t)	18.7/19.8/27.1/25.8	9.8/9.6/11.3/10.3	5.3/6.4/5.0/4.8
x ₄(t)	19.4/16.1/18.4/23.0	11.9/12.7/9.2/8.2	4.7/5.5/6.7/4.8

信号	d₁/Hz	d₂/Hz	d₃/Hz	d₄/Hz	d₅/Hz
x₁(t)	23.93	11.26	6.20	3.54	1.72
x ₂(t)	22.54	10.81	6.47	3.41	1.78
x ₃(t)	23.07	10.96	6.35	3.45	1.98
x ₄(t)	21.81	10.63	6.47	3.27	1.81

受试者	特征维度	基于NA-FMEMD的脑电信号分类方法	基于NA-MEMD的脑电信号分类方法
	m=1	70.0%	68.3%
	m=2	83.3%	78.3%
a	m=3	83.3%	85.0%
	m=4	85.0%	86.7%
	m=5	90.0%	81.7%
	m=1	71.7%	78.3%
	m=2	76.7%	75.0%
b	m=3	80.0%	73.3%
	m=4	78.3%	70.0%
	m=5	75.0%	68.3%
	m=1	60.0%	58.3%
	m=2	76.7%	80.0%
f	m=3	80.0%	78.3%
	m=4	80.0%	75.0%
	m=5	76.7%	73.3%
	m=1	68.3%	70.0%
	m=2	86.7%	75.0%
g	m=3	90.0%	86.7%
	m=4	90.0%	85.0%
	m=5	91.7%	88.3%

基于噪声辅助快速多维经验模式分解的运动想象脑电信号分类方法

Classification of motor imagery signals using noise-assisted fast multivariate empirical mode decomposition

在线阅读

PDF下载

可视化

摘要/Abstract

引用本文

使用本文

图/表 13

参考文献 29

相关文章 1

Metrics

推荐阅读 0

受试者	分类方法	基于NA-FMEMD的脑电信号分类方法	基于NA-MEMD的脑电信号分类方法
	SVM	90.0%	86.7%
a	CART	80.0%	85.0%
	KNN	83.3%	83.3%
	FLD	93.3%	80.0%
	SVM	80.0%	78.3%
b	CART	66.7%	65.0%
	KNN	68.3%	76.7%
	FLD	86.7%	83.3%