nav emailalert searchbtn searchbox tablepage yinyongbenwen piczone journalimg journalInfo journalinfonormal searchdiv searchzone qikanlogo popupnotification paper paperNew
Cross-session motor imagery decoding based on dual- branch contrastive alignment and domain adaptation
Email: zhouyueying@lcu.edu.cn
DOI: 10.19728/j.issn1672-6634.2026080004
Published:   2026-11-13
Publication Date:   2026-11-13
Online:   2026-11-13
Mobile reading
Abstract:

Motor imagery electroencephalography decoding is a core technical direction for non-invasive brain-computer interfaces. The performance degradation caused by limited cross-session generalization is a key bottleneck in practical applications, with its root in cross-session distribution shifts. To tackle this problem, this paper proposes a cross-session MI EEG decoding method based on dual-branch contrastive domain adaptation. The method applies Riemannian spatial filtering to extract spatial features, then extracts power spectral density and Pearson correlation coefficient features. Cross-view contrastive learning projects both features into a common latent space to achieve semantic alignment, so that the fused features encode both spectral and spatial information. The fused features are then fed into a domain-adversarial neural network. Through adversarial learning between the source and target domains, the model learns domain-invariant representations, mitigating the interference of cross-session distribution shifts on classification. Cross-session experiments are conducted on the SHU dataset. Results show that the proposed method achieves average classification accuracies of 59.76% and 59.91% on the two test sessions, respectively, outperforming several mainstream baseline methods. This work provides an effective direction for stable cross-session decoding in MI BCI.

References

[1] Wolpaw J R, Birbaumer N, McFarland D J, et al. Brain–computer interfaces for communication and control[J]. Clinical Neurophysiology, 2002, 113(6): 767-791.

[2] Jung T P, Makeig S, McKeown M J, et al. Imaging brain dynamics using independent component analysis[J]. Proceedings of the IEEE, 2001, 89(7): 1107-1122.

[3] 杨帮华. 运动想象脑机接口在神经系统疾病康复中的应用[J]. 康复学报, 2023, 33(6): 477-485.

[4] Saibene A, Ghaemi H, Dagdevir E. Deep learning in motor imagery EEG signal decoding: a systematic review[J]. Neurocomputing, 2024, 610: 128577.

[5] Pan L C, Wang K, Xu L C, et al. Riemannian geometric and ensemble learning for decoding cross-session motor imagery electroencephalography signals[J]. Journal of Neural Engineering, 2023, 20(6): 066011.

[6] Zheng L S, Ma Y, Lian P C, et al. A power spectrum pattern difference-based time-frequency sub-band selection method for MI-EEG classification[J]. IEEE Sensors Journal, 2022, 22(12): 11928-11939.

[7] Wimpff M, Döbler M, Yang B. Calibration-free online test-time adaptation for electroencephalography motor imagery decoding[C]//Proceedings of 12th International Winter Conference on Brain-Computer Interface. IEEE, 2024: 1-6.

[8] Wimpff M, Aristimunha B, Chevallier S, et al. Fine-tuning strategies for continual online EEG motor imagery decoding: insights from a large-scale longitudinal study[C]//Proceedings of 47th Annual International Conference of the IEEE Engineering in Medicine and Biology Society. IEEE, 2025: 1-7.

[9] Ramoser H, Muller-Gerking J, Pfurtscheller G. Optimal spatial filtering of single trial EEG during imagined hand movement[J]. IEEE Transactions on Rehabilitation Engineering, 2000, 8(4): 441-446.

[10] Blankertz B, Tomioka R, Lemm S, et al. Optimizing spatial filters for robust EEG single-trial analysis[J]. IEEE Signal Processing Magazine, 2008, 25(1): 41-56.

[11] Ang K K, Chin Z Y, Zhang H H, et al. Filter bank common spatial pattern (FBCSP) in brain-computer interface[C]//Proceedings of IEEE International Joint Conference on Neural Networks. IEEE, 2008: 2390-2397.

[12] Ang K K, Chin Z Y, Wang C C, et al. Filter bank common spatial pattern algorithm on BCI competition IV datasets 2a and 2b[J]. Frontiers in Neuroscience, 2012, 6: 39.

[13] 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): 5391-5420.

[14] Mane R, Robinson N, Vinod A P, et al. A multi-view CNN with novel variance layer for motor imagery brain computer interface[C]//Proceedings of 42nd Annual International Conference of the IEEE Engineering in Medicine & Biology Society. IEEE, 2020: 2950-2953.

[15] Miao Z Q, Zhao M R, Zhang X, et al. LMDA-Net: a lightweight multi-dimensional attention network for general EEG-based brain-computer interfaces and interpretability[J]. NeuroImage, 2023, 276: 120209.

[16] Fan C C, Yang B H, Li X O, et al. Temporal-frequency-phase feature classification using 3D-convolutional neural networks for motor imagery and movement[J]. Frontiers in Neuroscience, 2023, 17: 1250991.

[17] 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.

[18] Xu L C, Xu M P, Ke Y F, et al. Cross-dataset variability problem in EEG decoding with deep learning[J]. Frontiers in Human Neuroscience, 2020, 14: 103.

[19] Ganin Y, Ustinova E, Ajakan H, et al. Domain-adversarial training of neural networks[M]//Csurka G, Domain adaptation in computer vision applications. Cham: Springer, 2017: 189-209.

[20] She Q S, Chen T, Fang F, et al. Improved domain adaptation network based on Wasserstein distance for motor imagery EEG classification[J]. IEEE Transactions on Neural Systems and Rehabilitation Engineering, 2023, 31: 1137-1148.

[21] Hong X Z, Du C D, He H G. Adaptive domain alignment neural networks for cross-domain EEG emotion recognition[J]. IEEE Transactions on Affective Computing, 2025, 16(2): 903-914.

[22] 齐垒, 陈民铀, 张莉. 运动想象脑机接口的判别迁移特征学习与分类[J]. 重庆大学学报, 2024, 47(3): 86-95.

[23] 高云园, 薛云峰, 张聪睿, 等. 基于集成式张量域自适应的运动想象脑电分类[J]. 中国生物医学工程学报, 2024, 43(4): 399-407.

[24] 何敏, 罗天健. 多阶段分布适应的跨被试运动想象脑电解码模型[J]. 计算机应用, 2025, 45(12): 4064-4072.

[25] Wang Y H, Wang J X, Wang W Q, et al. Calibration-free transfer learning for EEG-based cross-subject motor imagery classification[C]//Proceedings of IEEE 19th International Conference on Automation Science and Engineering. IEEE, 2023: 1-6.

[26] Pan L C, Wang K, Huang Y Z, et al. Enhancing motor imagery EEG classification with a Riemannian geometry-based spatial filtering (RSF) method[J]. Neural Networks, 2025, 188: 107511.

[27] Ma J, Yang B H, Qiu W Z, et al. A large EEG dataset for studying cross-session variability in motor imagery brain-computer interface[J]. Scientific Data, 2022, 9: 531.

Basic Information:

DOI:10.19728/j.issn1672-6634.2026080004

China Classification Code:TN911.7;TP18;R318

Citation Information:

[1]MA Xiujin,ZHOU Yueying,WANG Lijun ,et al.Cross-session motor imagery decoding based on dual- branch contrastive alignment and domain adaptation[J].Journal of Liaocheng University (Natural Science Edition)().DOI:10.19728/j.issn1672-6634.2026080004.

Fund Information:

国家自然科学基金项目(62406131,62306139); 山东省自然科学基金项目(ZR2025MS1095); 山东省高等学校“青创团队计划”(2024KJN042)资助

Published:  

2026-11-13

Publication Date:  

2026-11-13

Online:  

2026-11-13

quote

GB/T 7714-2015
MLA
APA
Search Advanced Search