Algorithm Research Based on Multi-Feature Fusion of EEG Signals with Convolutional Neural Networks

doi:10.3778/j.issn.1002-8331.2212-0301

Abstract

Abstract: In order to address the issue of low classification accuracy in motor imagery of electroencephalogram (EEG) signals, a feature extraction algorithm based on sample entropy and common spatial pattern (CSP) feature fusion has been proposed. The algorithm initially performs wavelet packet decomposition on the raw EEG signal, selecting the components containing μ and β rhythms for reconstruction. Subsequently, the sample entropy and CSP features of the reconstructed signal are separately extracted. These two features are then fused to create a new feature vector which is recognized using a one-dimensional convolutional neural network designs in the paper, to obtain the classification result. The proposes method achieves a classification accuracy of 91.66% on the BCI Dataset III in 2003 and an average classification accuracy of 85.29% on the BCI Dataset A in 2008. Comparing with multi-feature fusion algorithms proposed in recent literature, the accuracy is improved by 7.96 percentage points.

Key words: electroencephalogram, motor imagery, wavelet packet transform, sample entropy, common spatial pattern, convolution neural network

摘要： 针对脑电信号（electroencephalogram，EEG）运动想象中单一特征无法多维表征信号中的信息导致的分类准确率不高的问题，提出一种基于样本熵和共空间模式特征融合的特征提取算法。算法先对原始脑电信号进行小波包分解，从中选择包含μ和β节律的分量进行重构，然后分别提取重构信号的样本熵和CSP（common spatial pattern，CSP）特征，将两者融合组成新的特征向量，使用所设计的一维卷积神经网络对其进行识别获得分类结果。所提方法在2003年BCI Dataset III中获得了91.66%的分类准确率，在2008年BCI Dataset A中获得了85.29%的平均分类准确率。与近年来文献中提出的多特征融合算法相比，准确率提高了7.96个百分点。

关键词: 脑电信号, 运动想象, 小波包重构, 样本熵, 共空间模式, 卷积神经网络

SONG Shilin, ZHANG Xuejun. Algorithm Research Based on Multi-Feature Fusion of EEG Signals with Convolutional Neural Networks[J]. Computer Engineering and Applications, 2024, 60(8): 148-155.

宋世林, 张学军. 脑电信号多特征融合与卷积神经网络算法研究[J]. 计算机工程与应用, 2024, 60(8): 148-155.

References

[1] AHMAD R F, MALIK A S, KAMEL N, et al. Discriminating the different human brain states with EEG signals using fractal dimension: a nonlinear approach[C]//Proceedings of the 2014 IEEE International Conference on Smart Instrumentation, Measurement and Applications, 2014.
[2] FAN M L, CHOU C A. Detecting abnormal pattern of epileptic seizures via temporal synchronization of EEG signals[J]. IEEE Transactions on Biomedical Engineering, 2019, 66(3): 601-608.
[3] JEONG J H, YU B W, LEE D H, et al. Classification of drowsiness levels based on a deep spatio-temporal convolutional bidirectional LSTM network using electroencephalography signals[J]. Brain Sciences, 2019, 9(12): 348.
[4] DENG X, ZHANG B, LIU K, et al. The classification of motor imagery EEG signals based on the time-frequency-spatial feature[C]//Proceedings of the 2021 IEEE 10th Data Driven Control and Learning Systems Conference, 2021: 6-11.
[5] 许学添, 蔡跃新. 基于图卷积网络的运动想象识别[J]. 计算机工程与应用, 2022, 58(4): 186-191.
XU X T, CAI Y X. Motor imagery recognition based on graph convolution network[J]. Computer Engineering and Applications, 2022, 58 (4): 186-191.
[6] POURALI H, OMRANPOUR H. CSP-Ph-PS: learning CSP-phase space and poincare sections based on evolutionary algorithm for EEG signals recognition[J]. Expert Systems with Applications, 2023, 211: 118621.
[7] ESPINOSA R, TALERO J, WEINSTEIN A. Effects of Tau and sampling frequency on the regularity analysis of ECG and EEG signals using ApEn and SampEn entropy estimators[J]. Entropy, 2020, 22(11): 1298.
[8] PAN C, SHI C, MU H, et al. EEG-based emotion recognition using logistic regression with Gaussian kernel and laplacian prior and investigation of critical frequency bands[J]. Applied Sciences, 2020, 10(5): 1619.
[9] 燕楠, 王珏, 魏娜, 等. 基于样本熵的注意力相关脑电特征信息提取与分类[J]. 西安交通大学学报, 2007(10):1237-1241.
YAN N, WANG J, WEI N, et al. Feature exaction and classification of attention related electroencephalographic signals based on sample entropy[J]. Journal of Xi’an Jiaotong University, 2007(10):1237-1241.
[10] AMIN S U, ALSULAIMAN M, MUHAMMAD G, et al. Multilevel weighted feature fusion using convolutional neural networks for EEG motor imagery classification[J]. IEEE Access, 2019, 7: 18940-18950.
[11] ECHTIOUI A, ZOUCH W, GHORBEL M, et al. Fusion convolutional neural network for multi-class motor imagery of EEG signals classification[C]//Proceedings of the 17th IEEE International Wireless Communications and Mobile Computing Conference, 2021.
[12] LI C, YANG H H, WU X, et al. Improving EEG-based motor imagery classification using hybrid neural network[C]//Proceedings of the IEEE 9th International Conference on Information, Communication and Networks, 2021.
[13] ZHANG L, HU J F, XIONG G L. Bearing fault diagnosis using a novel classifier ensemble based on lifting wavelet packet transforms and sample entropy[J]. Shock and Vibration, 2016, 2016(3):1-13.
[14] FRIZZI S, KAABI R, BOUCHOUICHA M, et al. Convolutional neural network for video fire and smoke detection[C]//Proceedings of the 42nd Annual Conference of the IEEE Industrial Electronics Society, 2016.
[15] WANG C, WU Y, WANG C, et al. MI-EEG classification using Shannon complex wavelet and convolutional neural networks[J]. Applied Soft Computing, 2022, 130: 109685.
[16] ZENG W, LI M Q, YUAN C Z, et al. Identification of epileptic seizures in EEG signals using time-scale decomposition (ITD), discrete wavelet transform (DWT), phase space reconstruction (PSR) and neural networks[J]. Artificial Intelligence Review, 2020, 53(4): 3059-3088.
[17] CIRUGEDA-ROLDAN E M, MOLINA-PICO A, CUESTA-FRAU D, et al. Comparative study between sample entropy and detrended fluctuation analysis performance on EEG records under data loss[C]//Proceedings of the 34th Annual International Conference of the IEEE Engineering-in-Medicine-and-Biology-Society, 2012: 4233-4236.
[18] 孔祥浩, 马琳, 薄洪健, 等. CNN与CSP相结合的脑电特征提取与识别方法研究[J]. 信号处理, 2018, 34(2): 164-173.
KONG X H, MA L, BO H J, et al. Research on EEG feature extraction and recognition method based on CNN and CSP[J]. Journal of Signal Processing, 2018, 34 (2): 164-173.
[19] MATSUO Y, LECUN Y, SAHANI M, et al. Deep learning, reinforcement learning, and world models[J]. Neural Networks, 2022, 152: 267-275.
[20] BLANKERTZ B, MULLER K R, CURIO G, et al. The BCI competition 2003: progress and perspectives in detection and discrimination of EEG single trials[J]. IEEE Transactions on Biomedical Engineering, 2004, 51(6): 1044-1051.
[21] FATOURECHI M, BASHASHATI A, WARD R K, et al. EMG and EOG artifacts in brain computer interface systems: a survey[J]. Clinical Neurophysiology, 2007, 118(3): 480-494.
[22] LIANG W, JIN J, DALY I, et al. Novel channel selection model based on graph convolutional network for motor imagery[J]. Cognitive Neurodynamics, 2022, 17:1283-1296.
[23] MA M Z, GUO L B, SU K F, et al. Classification of motor imagery EEG signals based on wavelet transform and sample entropy[C]//Proceedings of the 2nd IEEE Advanced Information Technology, Electronic and Automation Control Conference, 2017: 905-910.
[24] LI Y P, WANG Q, WANG T, et al. Feature extraction of EEG signals based on local mean decomposition and fuzzy entropy[J]. International Journal of Pattern Recognition and Artificial Intelligence, 2022, 61(6):4807-4820.
[25] 罗飞, 刘鹏飞, 罗元, 等. 多特征融合的运动想象脑电特征提取方法[J]. 计算机应用, 2020, 40(2): 616-620.
LUO F, LIU P F, LUO Y, et al. Motor imagery EEG feature extraction method based on multi-feature fusion[J]. Journal of Computer Applications, 2020, 40(2): 616-620.
[26] TABAR Y R, HALICI U. A novel deep learning approach for classification of EEG motor imagery signals[J]. Journal of Neural Engineering, 2017, 14(1): 016003.
[27] RAZA H, CHOWDHURY A, BHATTACHARYYA S, et al. Single-trial EEG classification with EEGNet and neural structured learning for improving BCI performance[C]//Proceedings of the International Joint Conference on Neural Networks (IJCNN) held as part of the IEEE World Congress on Computational Intelligence, 2020.