视觉分散自适应补偿的SSVEP多域协同解码研究

doi:10.19678/j.issn.1000-3428.0070595

摘要/Abstract

摘要： 基于稳态视觉诱发电位(SSVEP)的脑机接口(BCI)，由于个体差异和非目标刺激干扰问题面临分类性能瓶颈，且现有方法尚未深入探究视觉分散干扰与个体差异之间的量化关系。为此，本研究提出一种视觉分散自适应补偿的SSVEP多域协同解码算法，其核心包含视觉分散自适应标签平滑技术与多域联合解码模型两部分。首先，基于视觉拥挤效应理论，构建信号幅值与标签噪声的自适应量化模型，通过动态调节标签平滑强度实现个体视觉分散程度的量化表征。在缓解模型过拟合的同时，削弱非目标刺激的干扰影响，抑制个体差异的不利影响。进一步提出一种多域联合解码模型，先通过特征提取框架实现时频空域深度协同，再引入双向长短期记忆网络(Bi-LSTM)对时序全局依赖关系建模，形成兼具局部感受野与长程上下文感知能力的复合特征表达。在3个公开SSVEP数据集上以两个不同长度的时间窗（0.5s和1.0s）分别进行验证。实验表明，在所有实验设置下，该算法的平均识别准确率和平均信息传输率都更具优势。消融实验表明，视觉分散自适应补偿机制，在各实验设置下均能发挥作用，尤其是短时间窗口下，最高能提升18个百分点。表明该方法为神经解码中的个体适应性优化与时频特征融合提供了新的方法论框架。

Abstract: Steady-state visual evoked potential (SSVEP)-based brain-computer interfaces (BCIs) face classification performance bottlenecks due to individual variability and non-target stimulus interference. However, existing methodologies have not yet thoroughly investigated the quantitative correlation between visual distraction interference and inter-individual variability. To address this, we propose a multi-domain collaborative decoding algorithm with adaptive visual distraction compensation, comprising two core components: an adaptive label smoothing technique for visual distraction mitigation and a multi-domain joint decoding model. An adaptive quantification model based on visual crowding theory links signal amplitude to label noise. By dynamically adjusting label smoothing intensity, it quantifies individual distraction levels while reducing overfitting, non-target interference, and inter-subject variability. A multi-domain joint decoding model is proposed, establishing deep temporal-frequency-spatial synergy via hierarchical feature extraction and employing Bi-LSTM for global temporal dependencies. This hybrid architecture yields composite features with local receptive fields and long-range contextual awareness. Extensive experiments on three public SSVEP datasets under two time windows (0.5s and 1.0s) demonstrate the algorithm’s superiority. Results show consistent improvements in average classification accuracy and information transfer rate (ITR) across all settings. Ablation studies confirm the efficacy of the adaptive visual distraction compensation mechanism, particularly under short time windows (0.5s), where accuracy improves by up to 18 percentage points. This work provides a novel methodological framework for individualized adaptation and spatio-spectral-temporal feature fusion in neural decoding.

贾书婷, 温昕, 郝雁嵘, 曹锐. 视觉分散自适应补偿的SSVEP多域协同解码研究[J]. 计算机工程, doi: 10.19678/j.issn.1000-3428.0070595.

Jia Shuting, Wen Xin, Hao Yanrong, Cao Rui. Multi-Domain Collaborative Decoding of SSVEP with Adaptive Visual Distraction Compensation[J]. Computer Engineering, doi: 10.19678/j.issn.1000-3428.0070595.

参考文献

[1] ZHUANG M, WU Q, WAN F, et al. State-of-the-art non-invasive brain-computer interface for neural rehabilitation: A review[J]. Journal of Neurorestoratology, 2020, 8(1): 12-25.
[2] CHEN X, WANG Y, GAO S. EEG-based brain-computer interfaces: A review of recent progress in signal processing and applications[J]. IEEE Transactions on Cognitive and Developmental Systems, 2022, 14(3): 732-748.
[3] VEENA N, ANITHA N. A review of non-invasive BCI devices[J]. International Journal of Biomedical Engineering and Technology, 2020, 34(3): 205-233.
[4] Nakanishi M, Wang Y, Chen X. High-speed SSVEP-based BCI with filter bank canonical correlation analysis[J]. Journal of Neural Engineering, 2021, 18(4): 046028.
[5] 李明, 王伟, 张涛. 基于稳态视觉诱发电位的脑机接口系统设计与医疗应用[J]. 中国生物医学工程学报, 2021, 40(4): 432-439. LI M, WANG W, ZHANG T. Design and medical applications of SSVEP-based brain-computer interface systems[J]. Chinese Journal of Biomedical Engineering, 2021, 40(4): 432-439.
[6] 周晓燕, 刘洋, 高小榕. 可穿戴脑电设备在军事脑机接口中的最新进展 [J]. 仪器仪表学报 , 2023, 44(2): 156-165. ZHOU X, LIU Y, GAO X. Recent advances in wearable EEG devices for military brain-computer interfaces[J]. Chinese Journal of Scientific Instrument, 2023, 44(2): 156-165.
[7] MÜLLER-PUTZ G, et al. Benchmarking brain-computer interfaces outside the laboratory: A case study with SSVEP-based gaming[J]. Frontiers in Human Neuroscience, 2022, 16: 812345.
[8] ZHANG Y, et al. Enhancing SSVEP Detection Through Cortical Source Reconstruction[J]. NeuroImage, 2018, 183:897-908.
[9] WANG Y, JUNG T P. A Hybrid EEG-fNIRS Approach for Robust SSVEP Classification[J]. Journal of Neural Engineering, 2020, 17(3):036028.
[10] CHEN X, et al. Deep Learning-Based Spatial Filtering forHigh-Speed SSVEP-BCI[J]. IEEE Transactions on Biomedical Engineering, 2021, 68(4):1123-1132.
[11] ZHANG Q, WANG L, CHEN Z, et al. Cross-subject fluctuation analysis in SSVEP-based BCIs[J]. Journal of Neural Engineering, 2021, 18(4): 0460a3.
[12] CHEN X, LI H, WANG Y. Activation mechanisms of parieto-occipital cortex during SSVEP generation[J]. NeuroImage, 2021, 237: 118143.
[13] ROSENHOLTZ R. A unified model of crowding and visual search[J]. Vision Research, 2020, 173: 1-12.
[14] VAN DEN BERG R, ROERDINK J, CORNELISSEN F. FFT amplitude analysis of crowding effects in visual perception[J]. PLoS Computational Biology, 2020, 16(4): e1007842.
[15] NAKANISHI M, TANAKA T, WANG Y. Unsupervised frequency-recognition method of SSVEPs using a filter bank implementation of binary subband CCA[J]. IEEE Transactions on Biomedical Engineering, 2019, 66(3): 725-735.
[16] GUNEY O B, OBLOKULOV M, et al. A Deep Neural Network for SSVEP-Based Brain-Computer Interfaces[J]. Ieee Transactions on Biomedical Engineering, 2022, 69(2): 932-944.
[17] PAN Y D, LI N, et al. Short-time SSVEP data extension by a novel generative adversarial networks based framework[J]. arXiv preprint arXiv, 2023, 2301. 05599.
[18] MING D, LI R, GAO X. SSVEP-GAN: Subject-invariant feature learning via adversarial neural style transfer[J]. Medical Image Analysis, 2023, 88: 102856.
[19] PAN Y D, ZHANG Y, et al. A survey of deep learning-based classification methods for steady-state visual evoked potentials[J]. Brain-Apparatus. Communication: A Journal of Bacomics,2023,2(1): 2181102.
[20] PAN Y, CHEN J, ZHANG Y, et al. An efficient CNN-LSTM network with spectral normalization and label smoothing technologies for SSVEP frequency recognition[J]. Journal of Neural Engineering, 2022,19(5):102797.
[21] LIU Y, CHEN X, WANG Y. Deep frequency-domain residual network with multi-scale spectral pyramid for high-accuracy SSVEP decoding[J]. Journal of Neural Engineering, 2023, 20(3): 450-458.
[22] ZHANG S, AN D, LIU J, et al. Dynamic decomposition graph convolutional neural network for SSVEP-based brain-computer interface[J]. Neural networks: the official journal of the International Neural Network Society,2023,172:106075.
[23] 陈小刚, 李航, 王毅军. 基于时空卷积注意力网络的 SSVEP 解码方法研究[J]. 自动化学报, 2023, 49(5): 1021-1032. CHEN X G, LI H, WANG Y J. Spatio-temporal convolutional attention network for SSVEP decoding[J]. Acta Automatica Sinica, 2023, 49(5): 1021-1032.
[24] ZHANG Z, LI D D, ZHAO Y, et al. A flexible speller based on time-space frequency conversion SSVEP stimulation paradigm under dry electrode[J]. Frontiers In Computational Neuroscience, 2023,17: 1101726-1101726.
[25] DING W, SHAN J, FANG B, et al. Filter Bank Convolutional Neural Network for Short Time-Window Steady-State Visual Evoked Potential Classification[J]. Ieee Transactions on Neural Systems and Rehabilitation Engineering,2021,29:2615-2624.
[26] CHEN J N, SUN F C, et al. Attention-Based Multimodal tCNN for Classification of Steady-State Visual Evoked Potentials and Its Application to Gripper Control[J]. Ieee Transactions on Neural Networks and Learning Systems,2023,21(2):1124-1138.
[27] 明建国, 张伟, 刘洋. 混合神经网络在高速脑机接口中的应用[J]. 中国生物医学工程学报 , 2024, 43(2): 156-165.MING J G, ZHANG W, LIU Y. Application of hybrid neural networks in high-speed brain-computer interfaces[J]. Chinese Journal of Biomedical Engineering, 2024, 43(2): 156-165.
[28] RAVI A, BENI N H, MANUEL J, et al. Comparing user-dependent and user-independent training of CNN for SSVEP BCI[J]. Journal of Neural Engineering,2020,17(2):046080.
[29] CHEN J B, ZHANG Y S, PAN Y D, et al. A transformer-based deep neural network model for SSVEP classification[J]. Neural Networks,2023,164:521-534.
[30] NAKANISHI M, TANAKA T, WANG Y. DeepFreqNet: A multi-scale frequency-domain deep network for SSVEP classification[J]. IEEE Transactions on Biomedical Engineering, 2021, 68(12): 3565-3574.
[31] NAKANISHI M, WANG Y J, et al. A Comparison Study of Canonical Correlation Analysis Based Methods for Detecting Steady-State Visual Evoked Potentials[J]. Plos One,2015,10(10):18.
[32] WANG Y J, CHEN X G, GAO X R, et al. A Benchmark Dataset for SSVEP-Based Brain-Computer Interfaces[J]. Ieee Transactions on Neural Systems and Rehabilitation Engineering,2017,25(10):1746-1752.
[33] ZHU F, JIANG L, DONG G, et al. An open dataset for wearable SSVEP-based brain-computer interfaces[J]. Sensors, 2021, 21(3): 1256.
[34] LIU J W, WANG R M, YANG Y K, et al. Convolutional Transformer-Based Cross Subject Model for SSVEP-Based BCI Classification[J]. IEEE Journal of Biomedical and Health Informatics, 2024, 28(11): 6581-6593.
[35] QIN K, XU R, LI S R, et al. A Time-Local Weighted Transformation Recognition Framework for Steady State Visual Evoked Potentials Based Brain–Computer Interfaces[J]. IEEE Transactions on Neural Systems and Rehabilitation Engineering, 2024, 32(2): 1596-1605.

选择文件类型/文献管理软件名称

选择包含的内容