BCI(brain-computer interface)-MI(based Motor Imagery) training has been demonstrated to effectively enhance upper limb rehabilitation in stroke patients. However, traditional MI paradigms do not adequately address the diverse needs of patients regarding various joint movements during upper limb rehabilitation. A motor classification method based on SGE (spatial group-wise enhanced) multi-layer CNN (convolutional neural network) was proposed to improve the classification accuracy of upper limb movements. Six motor imagery paradigms targeting different joints of the unilateral upper limb were designed. A spatial grouping enhancement module was introduced to generate attention factors for each spatial location, adaptively adjusting the weight of EEG (electroencephalogram) data across different frequency bands. Features were then extracted using CNN, facilitating to the classification of unilateral upper limb movements. The method was experimentally validated on both public and self-collected datasets. The results indicate that the proposed method achieves higher accuracy for multi-class classification of unilateral upper limb movements, demonstrating its potential in BCI applications. It can be concluded that the proposed method provides new technical support for upper limb rehabilitation training in stroke patients.
GaoX, WangY, ChenX, et al. Interface, interaction, and intelligence in generalized brain-computer interfaces[J]. Trends in Cognitive Sciences, 2021, 25(8): 671-684.
BiWenlong, WeiXiao, TanCao, et al. Motor imagery classification based on PSO-SVM and its BCI control in lower limb exoskeleton[J]. Science Technology and Engineering, 2023, 23(16): 6952-6958.
ZhangYahui, WangFei, LiJinghong, et al. Semi-autonomous navigation control of intelligent wheelchair based on steady state visual evoked potential[J]. Robot, 2019, 41(5): 620-627, 636.
[8]
El-HassanH, El-MaaddawyT, Al-SallaminA, et al. Performance evaluation and microstructural characterization of GFRP bars in seawater-contaminated concrete[J]. Construction and Building Materials, 2017, 147: 66-78.
XiongTe, HuRonghua, ShaoHangfeng, et al. Control of upper limb rehabilitation robot based on steady-state visual evoked potential[J]. Science Technology and Engineering, 2021, 21(17): 7237-7242.
[11]
LiuD, XuX, LiD, et al. Intracranial brain-computer interface spelling using localized visual motion response[J]. NeuroImage, 2022, 258.DOI: 10.1016/j.neuroimage.2022.119363.
[12]
SuiM, OuyangY, JinH, et al. A soft-packaged and portable rehabilitation glove capable of closed-loop fine motorskills[J]. Nature Machine Intelligence, 2023, 5: 1149-1160.
[13]
TeoW P, EffieC. Is motor-imagery brain-computer interface feasible in stroke rehabilitation[J]. PM&R, 2014, 6(8): 723-728.
[14]
GernotR, MullerP, GertP. Control of an electrical prosthesis with an SSVEP-based BCI[J]. IEEE Transactions on Biomedical Engineering, 2007, 55(1): 361-364.
[15]
PolichJ. Updating P300: an integrative theory of P3a and P3b[J]. Clinical Neurophysiology, 2007, 118(10): 2128-2148.
[16]
GertP, ChristaN. Motor imagery and direct brain-computer communication[J]. Proceedings of the IEEE, 2001, 89(7): 1123-1134.
[17]
WagnerJ, MakeigS, GolaM, et al. Distinct β band oscillatory networks subserving motor and cognitive control during gait adaptation[J]. The Journal of Neuroscience: the Official Journal of the Society for Neuroscience, 2016, 36: 2212-2226.
[18]
ShimanF, López-LarrazE, Sarasola-SanzA, et al. Classification of different reaching movements from the same limb using EEG[J]. Journal of Neural Engineering, 2017, 14. DOI10.1088/1741-2552/aa70d2.
[19]
Harris-LoveM L, HarringtonR M. Non-invasive brain stimulation to enhance upper limb motor practice poststroke: a model for selection of cortical site[J]. Frontiers in Neurology, 2017, 8.DOI: 10.3389/fneur.2017.00224.
LiYuhua, LiuQian, WeiXin, et al. Classification of EEG signals based on convolutional neural networks[J]. Science Techno-logy and Engineering, 2020, 20(15): 6135-6140.
[22]
SchirrmeisterR T, SpringenbergJ T, FiedererL D, et al. Deep learning with convolutional neural networks for EEG decoding and visualization[J]. Human Brain Map, 2017, 38: 5391-5420.
RenLingling, WangLi, HuangXuewen, et al. Decoding motor and speech imagery EEG signals based on attention and multiscale model[J]. Science Technology and Engineering, 2022, 22(34): 15180-15187.
[25]
LawhernV, SolonJ, WaytowichN, et al. EEGNet: a compact convolutional neural network for EEG-based brain-computer interfaces[J]. Neural Engineering, 2018, 15(5). DOI: 10.1088/1741-2552/aace8c.
[26]
ChenX, TengH, Chen, Y, et al. Toward reliable signals decoding for electroencephalogram: a benchmark study to EEGNeX[J]. Biomedical Signal Processing and Control, 2024, 87.DOI: 10.1016/j.bspc.2023.105475.
[27]
OfnerP, SchwarzA, PereiraJ, et al. Upper limb movements can be decoded from the time-domain of low-frequency EEG[J]. PLoS ONE, 2017, 12(8). DOI: 10.1371/journal.pone.0182578.
[28]
DelormeA, ScottM. EEGLAB: an open source toolbox for analysis of single-trial EEG dynamics including independent component analysis[J]. Journal of Neuroscience Methods, 2004, 134(1): 9-21.
LiXiangzhe, WangDan, ZhangBaiwen, et all. A review on electroencephalogram based channel selection[J]. Journal of Biomedi-cal Engineering, 2024, 41(2): 398-405.
SongYaolian, YinXizhe, YangJun. Transformer based on temporal-spatial feature learning for motor imagery electroencephalogram signal decoding[J]. Journal of Nanjing Universit(Nature Science Edition), 2023, 59(2): 313-321.