Motor Imagery Classification Based on Plain Convolutional Neural Network and Linear Interpolation

doi:10.1007/s12204-022-2486-6

Abstract

Abstract: Deep learning has been applied for motor imagery electroencephalogram (MI-EEG) classification in brain-computer system to help people who suffer from serious neuromotor disorders. The inefficiency network and data shortage are the primary issues that the researchers face and need to solve. A novel MI-EEG classification method is proposed in this paper. A plain convolutional neural network (pCNN), which contains two convolution layers, is designed to extract the temporal-spatial information of MI-EEG, and a linear interpolation-based data augmentation (LIDA) method is introduced, by which any two unrepeated trials are randomly selected to generate a new data. Based on two publicly available brain-computer interface competition datasets, the experiments are conducted to confirm the structure of pCNN and optimize the parameters of pCNN and LIDA as well. The average classification accuracy values achieve 90.27% and 98.23%, and the average Kappa values are 0.805 and 0.965 respectively. The experiment results show the advantage of the proposed classification method in both accuracy and statistical consistency, compared with the existing methods.

Key words: motor imagery, classification, convolutional neural network, data augmentation, deep learning, braincomputer interface

摘要： 深度学习已应用于脑机接口系统中的运动想像脑电分类，以帮助患有严重运动神经障碍的人们。低效网络和数据短缺是研究者所面临和需要解决的主要问题。提出了一种新的运动想像脑电分类方法。设计一种含有两个卷积层的朴素卷积神经网络，以提取运动想像脑电信号的时空特征，并引入一种基于线性插值的数据增广方法，随机选择两个无重复的试验即可产生一个新数据。基于两个公共脑机接口竞赛数据集进行实验研究，确定了朴素卷积神经网络的结构，并优化了该网络及数据增广方法的参数；取得90.27%和98.23%的平均分类准确率，且平均卡帕值分别为0.85和0.965。实验结果表明：所提分类方法相对现有方法在识别精度和统计一致性方面均具有优势。

关键词: 运动想像，分类，卷积神经网络，数据增广，深度学习，脑机接口

CLC Number:

LI Mingai^{1, 2∗} (李明爱), WEI Lina¹ (魏丽娜). Motor Imagery Classification Based on Plain Convolutional Neural Network and Linear Interpolation[J]. J Shanghai Jiaotong Univ Sci, 2024, 29(6): 958-966.

References

[1] DAI M X, ZHENG D Z, NA R, et al. EEG classification of motor imagery using a novel deep learning framework [J]. Sensors, 2019, 19(3): 551.
[2] JOHNSON N N, CAREY J, EDELMAN B J, et al. Combined rTMS and virtual reality brain-computer interface training for motor recovery after stroke [J]. Journal of Neural Engineering, 2018, 15(1): 016009.
[3] TANG X L, LI W, LI X C, et al. Motor imagery EEG recognition based on conditional optimization empirical mode decomposition and multi-scale convolutional neural network [J]. Expert Systems with Applications, 2020, 149: 113285.
[4] ZHANG D L, YAO L N, CHEN K X, et al. Making sense of spatio-temporal preserving representations for EEG-based human intention recognition [J]. IEEE Transactions on Cybernetics, 2020, 50(7): 3033-3044.
[5] ZHAO X Q, ZHANG H M, ZHU G L, et al. A multibranch 3D convolutional neural network for EEGbased motor imagery classification [J]. IEEE Transactions on Neural Systems and Rehabilitation Engineering, 2019, 27(10): 2164-2177.
[6] DAI G H, ZHOU J, HUANG J H, et al. HS-CNN: a CNN with hybrid convolution scale for EEG motor imagery classification [J]. Journal of Neural Engineering,m 2020, 17(1): 016025.
[7] COLLAZOS-HUERTAS D F, ALVAREZ-MEZA A M,ACOSTA-MEDINA C D, et al. CNN-based framework using spatial dropping for enhanced interpretation of neural activity in motor imagery classification [J]. Brain Informatics, 2020, 7(1): 8.
[8] TANG X B, ZHAO J C, FU W L, et al. A novel classification algorithm for MI-EEG based on deep learning [C]//2019 IEEE 8th Joint International Information Technology and Artificial Intelligence Conference. Chongqing, China: IEEE, 2019: 606-611.
[9] GUBERT P H, COSTA M H, SILVA C D, et al. The performance impact of data augmentation in CSPbased motor-imagery systems for BCI applications [J]. Biomedical Signal Processing and Control, 2020, 62: 102152.
[10] LUO T J, ZHOU C L, CHAO F. Exploring spatialfrequency-sequential relationships for motor imagery classification with recurrent neural network [J]. BMC Bioinformatics, 2018, 19(1): 344.
[11] HARTMANN K G, SCHIRRMEISTER R T, BALL T. EEG-GAN: Generative adversarial networks for electroencephalograhic (EEG) brain signals [DB/OL]. (2018-06-05). https://arxiv.org/abs/1806.01875.
[12] SCHIRRMEISTER R T, SPRINGENBERG J T, FIEDERER L D J, et al. Deep learning with convolutional neural networks for EEG decoding and visualization [J]. Human Brain Mapping, 2017, 38(11):m 5391-5420.
[13] BCI Competition IV [EB/OL]. [2021-06-07]. https://www.bbci.de/competition/iv/.
[14] JIANG A M, SHANG J, LIU X F, et al. Efficient CSP algorithm with spatio-temporal filtering for motor imagery classification [J]. IEEE Transactions on Neural Systems and Rehabilitation Engineering, 2020, 28(4): 1006-1016.
[15] LI Z, YU Y. Improving EEG-based motor imagery classification with conditional Wasserstein GAN [C]//Proceedings of SPIE, 2020, 11584: 115841U.
[16] ZHENG Q Q, ZHU F Y, HENG P A. Robust support matrix machine for single trial EEG classification [J]. IEEE Transactions on Neural Systems and Rehabilitation Engineering, 2018, 26(3): 551-562.