您的位置：山东大学 -> 科技期刊社 -> 《山东大学学报（理学版）》

《山东大学学报(理学版)》 ›› 2025, Vol. 60 ›› Issue (7): 1-12.doi: 10.6040/j.issn.1671-9352.4.2024.126

• •    

EEG-MFNet:适用于脑电信号分析的轻量级多分支融合网络

叶晓雅1,2,3,王秀青1,2,3*,马海滨1,2,3,张诺飞1,2,3   

  1. 1.河北师范大学计算机与网络空间安全学院, 河北 石家庄 050024;2.河北省网络与信息安全重点实验室(河北师范大学), 河北 石家庄 050024;3.河北省供应链大数据分析与数据安全工程研究中心(河北师范大学), 河北 石家庄 050024
  • 发布日期:2025-07-01
  • 通讯作者: 王秀青(1970— ),女,教授,硕士生导师,博士,研究方向为脑电信号识别、脉冲神经网络、先进机器人技术等. ;E-mail:xqwang2013@163.com
  • 作者简介:叶晓雅(2000— ),女,硕士,研究方向为脑电信号识别、深度学习等. E-mail:13230051283@163.com*通信作者:王秀青(1970— ),女,教授,硕士生导师,博士,研究方向为脑电信号识别、脉冲神经网络、先进机器人技术等. E-mail:xqwang2013@163.com
  • 基金资助:
    国家自然科学基金资助项目(61673160);河北省自然科学基金资助项目(F2018205102);河北省高等学校科学技术研究重点项目(ZD2021063)

EEG-MFNet: a lightweight multi-branch fusion network for electroencephalogram signal analysis

YE Xiaoya1,2,3, WANG Xiuqing1,2,3*, MA Haibin1,2,3, ZHANG Nuofei1,2,3   

  1. 1. College of Computer and Cyber Security Hebei Normal University, Shijiazhuang 050024, Hebei, China;
    2. Hebei Provincial Key Laboratory of Network &
    Information Security(Hebei Normal University), Shijiazhuang 050024, Hebei, China;
    3. Hebei Provincial Engineering Research Center for Supply China Big Data Analytics &
    Data Security(Hebei Normal University), Shijiazhuang 050024, Hebei, China
  • Published:2025-07-01

PDF (PC)

8

摘要： 为解决脑电数据分辨率低、数据量不足以及被试者个体差异所导致的解码效率低下问题,提出适于脑电信号分析的轻量级多分支融合网络(multi-branch fusion network for electroencephalogram signal, EEG-MFNet)模型。通过多尺度时空卷积模块提取脑电数据的多层次时空特征,应用多尺度时间卷积提取更高级的时-空-频域特征,对输入分类器的特征数据应用滑动窗口,增强数据的有效特征。EEG-MFNet模型的平均分类准确率、标准差相比对比模型分别提升3.19%、22.86%以上,模型推理时间减少16.87%以上。实验结果表明所提方法提高运动想象脑电信号分类准确率,并增强模型的稳定性,提升了模型的训练效率,为基于运动想象的脑电信号分析提供更有效的解码方案。

关键词: 多尺度卷积, 滑动窗口, 运动想象, 脑机接口, 脑电信号

Abstract: In order to solve the problems of decoding efficiency caused by low resolution, insufficient data volume and individual differences of subjects, multi-branch fusion network for electroencephalogram signal(EEG-MFNet)model suitable for EEG signal analysis is proposed. Multi-level spatiotemporal features of EEG data are extracted through multi-scale spatiotemporal convolutional modules, and further applied by multi-scale temporal convolution to extract more advanced time-space-frequency domain features. Applying a sliding window to the feature data of the input classifier significantly enhances the effective features of the data. The average classification accuracy and standard deviation of the EEG-MFNet model are improved by more than 3.19% and 22.86% compared with the comparison model, respectively. Model inference time is reduced by more than 16.87%. The experimental results show that the proposed method improves the stability of the model and significantly improves the training efficiency. This work provides a more efficient decoding scheme for EEG signal analysis based on motor imagery.

Key words: multi-scale convolution, slide window, motor imagery, brain-computer interface, electroencephalogram signal

中图分类号: 

  • TP391
[1] JIAO Runhai, LI Chengyang, XUN Gangyi, et al. A context-aware multi-event identification method for nonintrusive load monitoring[J]. IEEE Transactions on Consumer Electronics, 2023, 69(2):194-204.
[2] ZHANG Yu, LI Penghai, CHENG Longlong, et al. Attention-based multiscale spatial-temporal convolutional network for motor imagery EEG decoding[J]. IEEE Transactions on Consumer Electronics, 2024, 70(1):2423-2434.
[3] AHMED I, JEON G, PICCIALLI F. From artificial intelligence to explainable artificial intelligence in industry 4.0: a survey on what, how, and where[J]. IEEE Transactions on Industrial Informatics, 2022, 18(8):5031-5042.
[4] WILLIAMS S C, HORSFALL H L, FUNNELL J P, et al. Neurosurgical team acceptability of brain-computer interfaces: a two-stage international cross-sectional survey[J]. World Neurosurgery, 2022, 164:884-898.
[5] LI Hongli, LIU Haoyu, CHEN Hongyu, et al. Multi-scale feature extraction and classification of motor imagery electroencephalography based on time series data enhancement[J]. Journal of Biomedical Engineering, 2023, 40(3):418-425.
[6] CHUNDURI V, AOUNDNI Y, KHAN S, et al. Multi-scale spatiotemporal attention network for neuron based motor imagery EEG classification[J]. Journal of Neuroscience Methods, 2024, 406:110-128.
[7] GAO Dongrui, YANG Wen, LI Pengrui, et al. A multiscale feature fusion network based on attention mechanism for motor imagery EEG decoding[J]. Applied Soft Computing, 2024, 151:111-129.
[8] LI Hongli, CHEN Hongyu, JIA Ziyu, et al. A parallel multi-scale time-frequency block convolutional neural network based on channel attention module for motor imagery classification[J]. Biomedical Signal Processing and Control, 2023, 79:104066.
[9] LAWHERN V J, SOLON A J, WAYTOWICH N R, et al. EEGNet: a compact convolutional neural network for EEG-based brain-computer interfaces[J]. Journal of neural engineering, 2018, 15(5):056013.
[10] ZHANG Yu, WANG Yu, ZHOU Guoxu, et al. Multi-kernel extreme learning machine for EEG classification in brain-computer interfaces[J]. Expert Systems with Applications, 2018, 96:302-310.
[11] ANG K K, CHIN Z Y, ZHANG H, et al. Filter bank common spatial pattern(FBCSP)in brain-computer interface[C] //2008 IEEE International Joint Conference on Neural Networks, Hong Kong: IEEE, 2390-2397.
[12] 吕璐,程虎,朱鸿泰,等. 基于深度学习的目标检测研究与应用综述[J]. 电子与封装,2022,22(1):72-80. LV Lu, CHENG Hu, ZHU Hongtai, et al. Review of research and application of target detection based on deep learning[J]. Electrons and Packaging, 2022, 22(1):72-80.
[13] INGOLFSSON T M, HERSCHE M, WANG X, et al. EEG-TCNet: an accurate temporal convolutional network for embedded motor-imagery brain-machine interfaces[C] //2020 IEEE International Conference on Systems, Man, and Cybernetics(SMC). Toronto: IEEE, 2020:2958-2965.
[14] ALTAHERI H, MUHAMMAD G, ALSULAIMAN M. Dynamic convolution with multilevel attention for EEG-based motor imagery decoding[J]. IEEE Internet of Things Journal, 2023, 10(21):18579-18588.
[15] ALTUWAIJRI G, MUHAMMAD G. A multibranch of convolutional neural network models for electroencephalogram-based motor imagery classification[J]. Biosensors, 2022, 12(1):22.
[16] 王蒙昊,方慧娟,龚亨翔,等. 应用多尺度混合卷积网络的脑电信号特征提取与识别[J]. 华侨大学学报(自然科学版),2023,44(5):628-635. WANG Menghao, FANG Huijuan, GONG Hengxiang, et al. Feature extraction and identification of EEG signals using multi-scale hybrid convolutional networks[J]. Journal of Huaqiao University(Natural Science Edition), 2023, 44(5):628-635.
[17] TANG Xianlun, YANG Caiquan, SUN Xia, et al. Motor imagery EEG decoding based on multi-scale hybrid networks and feature enhancement[J]. IEEE Transactions on Neural Systems and Rehabilitation Engineering, 2023, 31:1208-1218.
[18] GUENNEC A L, MALINOWSKI S, TAVENARD R. Data augmentation for time series classification using convolutional neural networks[C] //European Conference on Principles of Data Mining and Knowledge Discovery. Riva del Garda, Italy: IEEE, 2016.
[19] KIM S K, KIRCHNER E A, STEFES A, et al. Intrinsic interactive reinforcement learning-using error-related potentials for real world human-robot interaction[J]. Scientific Reports, 2017, 7(1):1-16.
[20] LOTTE F. Signal processing approaches to minimize or suppress calibration time in oscillatory activity-based brain-computer interfaces[J]. Proceedings of the IEEE, 2015, 103(6):871-890.
[21] WANG Fang, ZHONG Shenhua, PENG Jianfen, et al. Data augmentation for EEG-based emotion recognition with deep convolutional neural networks[C] //MultiMedia Modeling: 24th International Conference, MMM 2018. Tokyo: Springer, 2018.
[22] LI Youjun, HUANG Jiajin, ZHOU Haiyan, et al. Human emotion recognition with electroencephalographic multidimensional features by hybrid deep neural networks[J]. Applied Sciences, 2017, 7(10):1060.
[23] BRUNNER C, LEEB R, MULLER-PUTZ G et al. BCI competition 2008-Graz data set A[DB/OL].(2008-07-03)[2014-01-17]. https://ieee-dataport.org/documents/bci-competition-2008.
[24] ZHANG C, KIM Y K, ESKANDARIAN A. EEG-inception: an accurate and robust end-to-end neural network for EEG-based motor imagery classification[J]. Journal of Neural Engineering, 2021, 18(4):046014.
[25] SALAMI A, ANDREU-PEREZ J, GILLMEISTER H. EEG-ITNet: an explainable inception temporal convolutional network for motor imagery classification[J]. IEEE Access, 2022, 10:36672-36685.
[26] 刘近贞,叶方方,熊慧. 基于卷积神经网络的多类运动想象脑电信号识别[J]. 浙江大学学报(工学版),2021,55(11):2054-2066. LIU Jinzhen, YE Fangfang, XIONG Hui. Multi-class motor imagery EEG signal recognition based on convolutional neural networks[J]. Journal of Zhejiang University(Engineering Edition), 2021, 55(11):2054-2066.
[27] KIRANYAZ S, AVCI O, ABDELJABER O, et al. 1D convolutional neural networks and applications: a survey[J]. Mechanical Systems and Signal Processing, 2021, 151:107398.
[28] SCHIRRMEISTER R T, SPRINGENBERG J T, FIEDERER L D J, et al. Deep learning with convolutional neural networks for EEG decoding and visualization[C] //2017 IEEE Signal Processing in Medicine and Biology Symposium(SPMB). Los Angeles: IEEE, 2017:1-7.
[29] YANG Jun, MA Zhengmin, SHEN Tao. Multi-time and multi-band CSP motor imagery EEG feature classification algorithm [J]. Applied Sciences, 2021, 11(21):10294.
[30] ALTUWAIJRI G A, MUHAMMAD G. Electroencephalogram-based motor imagery signals classification using a multi-branch convolutional neural network model with attention blocks[J]. Bioengineering(Basel), 2022, 9(7):323.
[31] MA Weifeng, XUE Haojie, SUN Xiaoyong, et al. A novel multi-branch hybrid neural network for motor imagery EEG signal classification [J]. Biomedical Signal Processing and Control, 2022, 77:103718.
[1] 毋泽南,田立勤,王志刚. 一种结合滑动窗口和推荐信任的用户行为信任评估[J]. 《山东大学学报(理学版)》, 2019, 54(1): 53-59.
[2] 黄崇争,吴元锡,陈 红 . 数据流中一种有效的当前频繁序列挖掘方法[J]. J4, 2007, 42(11): 37-39 .
Viewed
Full text


Abstract

Cited
  Shared   
  Discussed   
No Suggested Reading articles found!
版权所有 © 2018 《山东大学学报(理学版)》编辑部　
地址: 山东省济南市山大南路27号山东大学中心校区（250100）　电话: +86-0531-88366917　E-mail: xblxb@sdu.edu.cn
本系统由北京玛格泰克科技发展有限公司设计开发