《计算机应用》唯一官方网站 ›› 2025, Vol. 45 ›› Issue (6): 1849-1857.DOI: 10.11772/j.issn.1001-9081.2024060794
• 人工智能 • 上一篇
周天彤1,2, 郑妍琪1,2, 魏韬1, 戴亚康3, 邹凌2,4()
收稿日期:
2024-06-13
修回日期:
2024-10-12
接受日期:
2024-10-18
发布日期:
2024-11-01
出版日期:
2025-06-10
通讯作者:
邹凌
作者简介:
周天彤(1972—),男,江苏常州人,副教授,硕士,主要研究方向:人机交互基金资助:
Tiantong ZHOU1,2, Yanqi ZHENG1,2, Tao WEI1, Yakang DAI3, Ling ZOU2,4()
Received:
2024-06-13
Revised:
2024-10-12
Accepted:
2024-10-18
Online:
2024-11-01
Published:
2025-06-10
Contact:
Ling ZOU
About author:
ZHOU Tiantong, born in 1972, M. S., associate professor. His research interests include human-computer interaction.Supported by:
摘要:
针对认知负荷识别模型存在过于依赖手动特征提取、忽视脑电图(EEG)信号的空间信息和无法有效学习图结构数据的问题,提出一种融合变分图自编码器(VGAE)与局部-全局图网络(VLGGNet)的认知负荷EEG识别模型。该模型由时间学习模块和图形学习模块这2个部分组成。首先,使用时间学习模块通过多尺度时间卷积捕捉EEG信号的动态频率表示,并通过空间与通道重建卷积(SCConv)和1
中图分类号:
周天彤, 郑妍琪, 魏韬, 戴亚康, 邹凌. 融合变分图自编码器与局部-全局图网络的认知负荷脑电识别模型[J]. 计算机应用, 2025, 45(6): 1849-1857.
Tiantong ZHOU, Yanqi ZHENG, Tao WEI, Yakang DAI, Ling ZOU. Cognitive load EEG recognition model integrating variational graph autoencoder and local-global graph network[J]. Journal of Computer Applications, 2025, 45(6): 1849-1857.
模型 | MAT数据集 | STEW 数据集 | ||
---|---|---|---|---|
mAcc | mF1 | mAcc | mF1 | |
DeepConvNet | 50.79*** | 48.03*** | 92.85*** | 92.20*** |
R2G-STNN | 59.81*** | 61.29*** | 95.68*** | 95.68*** |
EEGNet | 58.15*** | 57.37*** | 96.76*** | 96.69*** |
MTFB_CNN | 68.06*** | 72.11*** | 93.67*** | 94.12*** |
ShallowConvNet | 86.94*** | 87.64*** | 96.43*** | 96.44*** |
TSception | 93.10*** | 92.50*** | 98.76 | 98.78 |
LGGNet | 93.24*** | 93.45*** | 98.28** | 98.27** |
VLGGNet | 97.31 | 97.31 | 98.76 | 98.77 |
表1 不同数据集上8个模型的mAcc和mF1的对比 (%)
Tab. 1 Comparison of mAcc and mF1 of eight models on different datasets
模型 | MAT数据集 | STEW 数据集 | ||
---|---|---|---|---|
mAcc | mF1 | mAcc | mF1 | |
DeepConvNet | 50.79*** | 48.03*** | 92.85*** | 92.20*** |
R2G-STNN | 59.81*** | 61.29*** | 95.68*** | 95.68*** |
EEGNet | 58.15*** | 57.37*** | 96.76*** | 96.69*** |
MTFB_CNN | 68.06*** | 72.11*** | 93.67*** | 94.12*** |
ShallowConvNet | 86.94*** | 87.64*** | 96.43*** | 96.44*** |
TSception | 93.10*** | 92.50*** | 98.76 | 98.78 |
LGGNet | 93.24*** | 93.45*** | 98.28** | 98.27** |
VLGGNet | 97.31 | 97.31 | 98.76 | 98.77 |
多尺度 时间卷积 | 特征融合 模块 | 全局特征 提取 | MAT数据集 | STEW 数据集 | ||
---|---|---|---|---|---|---|
mAcc | mF1 | mAcc | mF1 | |||
√ | √ | 96.02 | 96.07 | 97.38 | 97.43 | |
√ | √ | 97.13 | 97.16 | 98.25 | 98.24 | |
√ | √ | 81.34 | 82.25 | 96.94 | 96.91 | |
√ | √ | √ | 97.31 | 97.31 | 98.76 | 98.77 |
表2 消融研究结果 (%)
Tab. 2 Ablation study results
多尺度 时间卷积 | 特征融合 模块 | 全局特征 提取 | MAT数据集 | STEW 数据集 | ||
---|---|---|---|---|---|---|
mAcc | mF1 | mAcc | mF1 | |||
√ | √ | 96.02 | 96.07 | 97.38 | 97.43 | |
√ | √ | 97.13 | 97.16 | 98.25 | 98.24 | |
√ | √ | 81.34 | 82.25 | 96.94 | 96.91 | |
√ | √ | √ | 97.31 | 97.31 | 98.76 | 98.77 |
不同尺度系数组合 | MAT数据集 | STEW 数据集 | ||
---|---|---|---|---|
mAcc | mF1 | mAcc | mF1 | |
[0.125,0.250,0.375] | 94.12 | 94.27 | 98.21 | 98.36 |
[0.250,0.375,0.500] | 91.16 | 91.63 | 98.01 | 98.06 |
[0.375,0.500,0.625] | 96.57 | 96.40 | 98.43 | 98.43 |
[0.500,0.625,0.750] | 92.08 | 92.22 | 98.47 | 98.48 |
[0.250,0.500,0.750] | 96.90 | 96.99 | 98.61 | 98.61 |
[0.500,0.750,1.000] | 95.32 | 95.23 | 98.18 | 98.26 |
[0.125,0.250,0.500] | 97.31 | 97.31 | 98.76 | 98.77 |
表3 不同尺度时间卷积组合在不同数据集上的性能比较 (%)
Tab. 3 Performance comparison of different scale temporal convolution combinations on different datasets
不同尺度系数组合 | MAT数据集 | STEW 数据集 | ||
---|---|---|---|---|
mAcc | mF1 | mAcc | mF1 | |
[0.125,0.250,0.375] | 94.12 | 94.27 | 98.21 | 98.36 |
[0.250,0.375,0.500] | 91.16 | 91.63 | 98.01 | 98.06 |
[0.375,0.500,0.625] | 96.57 | 96.40 | 98.43 | 98.43 |
[0.500,0.625,0.750] | 92.08 | 92.22 | 98.47 | 98.48 |
[0.250,0.500,0.750] | 96.90 | 96.99 | 98.61 | 98.61 |
[0.500,0.750,1.000] | 95.32 | 95.23 | 98.18 | 98.26 |
[0.125,0.250,0.500] | 97.31 | 97.31 | 98.76 | 98.77 |
图构建方法 | MAT数据集 | STEW数据集 | ||
---|---|---|---|---|
mAcc | mF1 | mAcc | mF1 | |
VGAE | 97.31 | 97.31 | 98.76 | 98.77 |
计算相似性 | 96.80 | 96.72 | 98.53 | 98.54 |
表4 VGAE与其他图构建方法在不同数据集上的性能比较 (%)
Tab. 4 Performance comparison of VGAE and other graph reconstruction methods on different datasets
图构建方法 | MAT数据集 | STEW数据集 | ||
---|---|---|---|---|
mAcc | mF1 | mAcc | mF1 | |
VGAE | 97.31 | 97.31 | 98.76 | 98.77 |
计算相似性 | 96.80 | 96.72 | 98.53 | 98.54 |
1 | GEVINS A, SMITH M E. Neurophysiological measures of cognitive workload during human-computer interaction [J]. Theoretical Issues in Ergonomics Science, 2003, 4(1/2): 113-131. |
2 | BERNHARDT K A, POLTAVSKI D, PETROS T, et al. The effects of dynamic workload and experience on commercially available EEG cognitive state metrics in a high-fidelity air traffic control environment [J]. Applied Ergonomics, 2019, 77: 83-91. |
3 | BADDELEY A. Working memory: the interface between memory and cognition [J]. Journal of Cognitive Neuroscience, 1992, 4(3): 281-288. |
4 | ANTONENKO P, PAAS F, GRABNER R, et al. Using electroencephalography to measure cognitive load[J]. Educational Psychology Review, 2010, 22(4): 425-438. |
5 | LIU R, QI S, HAO S, et al. Drivers’ workload electroencephalogram characteristics in cognitive tasks based on improved multiscale sample entropy[J]. IEEE Access, 2023, 11: 42180-42190. |
6 | BANNERT M. Managing cognitive load — recent trends in cognitive load theory [J]. Learning and Instruction, 2002, 12(1): 139-146. |
7 | HÄMÄLÄINEN M, HARI R, ILMONIEMI R J, et al. Magnetoencephalography — theory, instrumentation, and applications to noninvasive studies of the working human brain[J]. Reviews of Modern Physics, 1993, 65(2): 413-497. |
8 | VANDEWALLE G, ARCHER S N, WUILLAUME C, et al. Effects of light on cognitive brain responses depend on circadian phase and sleep homeostasis[J]. Journal of Biological Rhythms, 2011, 26(3): 249-259. |
9 | WANG H, XU L, BEZERIANOS A, et al. Linking attention-based multiscale CNN with dynamical GCN for driving fatigue detection[J]. IEEE Transactions on Instrumentation and Measurement, 2021, 70: No.2504811. |
10 | 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. |
11 | ZHANG K, YE Z, AI Q, et al. GNN4EEG: a benchmark and toolkit for electroencephalography classification with graph neural network[C]// Companion of the 2024 ACM International Joint Conference on Pervasive and Ubiquitous Computing. New York: ACM, 2024: 612-617. |
12 | CHAO H, CAO Y, LIU Y. Multi-channel EEG emotion recognition through residual graph attention neural network[J]. Frontiers in Neuroscience, 2023, 17: No.1135850. |
13 | GRAÑA M, MORAIS-QUILEZ I. A review of graph neural networks for electroencephalography data analysis[J]. Neurocomputing, 2023, 562: No.126901. |
14 | MUELLER T T, STARCK S, FEINER L F, et al. Extended graph assessment metrics for regression and weighted graphs[C]// Proceedings of the 2023 International Conference on Medical Image Computing and Computer-Assisted Intervention, LNCS 14373. Cham: Springer, 2024: 14-26. |
15 | GENG X, LI Y, WANG L, et al. Spatiotemporal multi-graph convolution network for ride-hailing demand forecasting[C]// Proceedings of the 33rd AAAI Conference on Artificial Intelligence. Palo Alto: AAAI Press, 2019: 3656-3663. |
16 | DING Y, ROBINSON N, TONG C, et al. LGGNet: learning from local-global-graph representations for brain-computer interface[J]. IEEE Transactions on Neural Networks and Learning Systems, 2024, 35(7): 9773-9786. |
17 | ZYMA I, TUKAEV S, SELEZNOV I, et al. Electroencephalograms during mental arithmetic task performance[J]. Data, 2019, 4(1): No.14. |
18 | LIM W L, SOURINA O, WANG L P. STEW: simultaneous task EEG workload data set [J]. IEEE Transactions on Neural Systems and Rehabilitation Engineering, 2018, 26(11): 2106-2114. |
19 | ZHU G, ZONG F, ZHANG H, et al. Cognitive load during multitasking can be accurately assessed based on single channel electroencephalography using graph methods[J]. IEEE Access, 2021, 9: 33102-33109. |
20 | YEDUKONDALU J, SHARMA L D. Cognitive load detection using circulant singular spectrum analysis and Binary Harris Hawks Optimization based feature selection [J]. Biomedical Signal Processing and Control, 2023, 79(Pt 1): No.104006. |
21 | 吴选昆,颜延,贾振华,等. 基于脑电功能连接拓扑表征的心算任务分类[J]. 计算机应用研究, 2022, 39(2): 356-360. |
WU X K, YAN Y, JIA Z H, et al. Mental arithmetic task classification based on topological representation of EEG-based functional connectivity[J]. Application Research of Computers, 2022, 39(2): 356-360. | |
22 | GUAN K, ZHANG Z, CHAI X, et al. EEG based dynamic functional connectivity analysis in mental workload tasks with different types of information[J]. IEEE Transactions on Neural Systems and Rehabilitation Engineering, 2022, 30: 632-642. |
23 | GANGULY B, CHATTERJEE A, MEHDI W, et al. EEG based mental arithmetic task classification using a stacked long short term memory network for brain-computer interfacing[C]// Proceedings of the 2020 IEEE International Conference on VLSI Device, Circuit and System. Piscataway: IEEE, 2020: 89-94. |
24 | MALVIYA L, MAL S. A novel technique for stress detection from EEG signal using hybrid deep learning model[J]. Neural Computing and Applications, 2022, 34(22): 19819-19830. |
25 | VARSHNEY A, GHOSH S K, PADHY S, et al. Automated classification of mental arithmetic tasks using recurrent neural network and entropy features obtained from multi-channel EEG signals [J]. Electronics, 2021, 10(9): No.1079. |
26 | JANG S, MOON S E, LEE J S. EEG-based video identification using graph signal modeling and graph convolutional neural network[C]// Proceedings of the 2018 IEEE International Conference on Acoustics, Speech and Signal Processing. Piscataway: IEEE, 2018: 3066-3070. |
27 | SONG T, ZHENG W, LIU S, et al. Graph-embedded convolutional neural network for image-based EEG emotion recognition[J]. IEEE Transactions on Emerging Topics in Computing, 2022, 10(3): 1399-1413. |
28 | ZHONG P, WANG D, MIAO C. EEG-based emotion recognition using regularized graph neural networks [J]. IEEE Transactions on Affective Computing, 2022, 13(3): 1290-1301. |
29 | LI Y, ZHENG W, WANG L, et al. From regional to global brain: a novel hierarchical spatial-temporal neural network model for EEG emotion recognition[J]. IEEE Transactions on Affective Computing, 2022, 13(2): 568-578. |
30 | DELVIGNE V, WANNOUS H, DUTOIT T, et al. PhyDAA: physiological dataset assessing attention [J]. IEEE Transactions on Circuits and Systems for Video Technology, 2022, 32(5): 2612-2623. |
31 | GILAKJANI S S, OSMAN H AL. A graph neural network for EEG-based emotion recognition with contrastive learning and generative adversarial neural network data augmentation [J]. IEEE Access, 2024, 12:113-130. |
32 | 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): No.056013. |
33 | CHEN C S, CHIEN T S, LEE P L, et al. Prefrontal brain electrical activity and cognitive load analysis using a non-linear and non-stationary approach [J]. IEEE Access, 2020, 8: 211115-211124. |
34 | DING Y, ROBINSON N, ZENG Q, et al. TSception: a deep learning framework for emotion detection using EEG[C]// Proceedings of the 2020 International Joint Conference on Neural Networks. Piscataway: IEEE, 2020: 1-7. |
35 | IOFFE S, SZEGEDY C. Batch normalization: accelerating deep network training by reducing internal covariate shift[C]// Proceedings of the 32nd International Conference on Machine Learning. New York: JMLR.org, 2015: 448-456. |
36 | LI J, WEN Y, HE L. SCConv: spatial and channel reconstruction convolution for feature redundancy[C]// Proceedings of the 2023 IEEE/CVF Conference on Computer Vision and Pattern Recognition. Piscataway: IEEE, 2023: 6153-6162. |
37 | WU Y, HE K. Group normalization[C]// Proceedings of the 2018 European Conference on Computer Vision, LNCS 11217. Cham: Springer, 2018: 3-19. |
38 | POWER J D, COHEN A L, NELSON S M, et al. Functional network organization of the human brain[J]. Neuron, 2011, 72(4): 665-678. |
39 | FRIEDERICI A D, CHOMSKY N, BERWICK R C, et al. Language, mind and brain [J]. Nature Human Behaviour, 2017, 1(10): 713-722. |
40 | KOBER H, BARRETT L F, JOSEPH J, et al. Functional grouping and cortical-subcortical interactions in emotion: a meta-analysis of neuroimaging studies [J]. NeuroImage, 2008, 42(2): 998-1031. |
41 | ALARCÃO S M, FONSECA M J. Emotions recognition using EEG signals: a survey [J]. IEEE Transactions on Affective Computing, 2019, 10(3): 374-393. |
42 | SALVADOR R, SUCKLING J, COLEMAN M R, et al. Neurophysiological architecture of functional magnetic resonance images of human brain[J]. Cerebral Cortex, 2005, 15(9): 1332-1342. |
43 | KIPF T N, WELLING M. Variational graph auto-encoders[EB/OL]. [2024-11-21].. |
44 | JIANG Y, HUANG C, HUANG L. Adaptive graph contrastive learning for recommendation[C]// Proceedings of the 29th ACM SIGKDD Conference on Knowledge Discovery and Data Mining. New York: ACM, 2023: 4252-4261. |
45 | LI H, CHEN H, JIA Z, 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(Pt 1): No.104066. |
46 | VARMA S, SIMON R. Bias in error estimation when using cross-validation for model selection [J]. BMC Bioinformatics, 2006, 7: No.91. |
47 | KOELSTRA S, MUHL C, SOLEYMANI M, et al. DEAP: a database for emotion analysis; using physiological signals[J]. IEEE Transactions on Affective Computing, 2012, 3(1): 18-31. |
48 | SIMONYAN K, VEDALDI A, ZISSERMAN A. Deep inside convolutional networks: visualising image classification models and saliency maps [EB/OL]. [2024-12-20].. |
49 | ABADI H E, MORIDANI M K, MIRZAKHANI M. Mental arithmetic task detection using geometric features extraction of EEG signal based on machine learning[J]. Bratislava Medical Journal, 2022, 123(6): 408-421. |
50 | TOMASI D, GOLDSTEIN R Z, TELANG F, et al. Widespread disruption in brain activation patterns to a working memory task during cocaine abstinence[J]. Brain Research, 2007, 1171: 83-92. |
[1] | 晁浩, 封舒琪, 刘永利. 脑电情感识别中多上下文向量优化的卷积递归神经网络[J]. 《计算机应用》唯一官方网站, 2024, 44(7): 2041-2046. |
[2] | 时旺军, 王晶, 宁晓军, 林友芳. 小样本场景下的元迁移学习睡眠分期模型[J]. 《计算机应用》唯一官方网站, 2024, 44(5): 1445-1451. |
[3] | 曹铉, 罗天健. 运动想象脑电信号的跨被试动态多域对抗学习方法[J]. 《计算机应用》唯一官方网站, 2024, 44(2): 645-653. |
[4] | 杨思琪, 罗天健, 严宣辉, 杨光局. 运动想象脑电图的空域特征迁移核学习方法[J]. 《计算机应用》唯一官方网站, 2024, 44(11): 3354-3363. |
[5] | 杨淑莹, 国海铭, 李欣. 基于通道选择和多维特征融合的脑电信号分类[J]. 《计算机应用》唯一官方网站, 2023, 43(11): 3418-3427. |
[6] | 张军鹏, 施玉杰, 蒋睿, 董静静, 邱昌建. 基于脑电信号的认知功能障碍识别与分类进展综述[J]. 《计算机应用》唯一官方网站, 2023, 43(10): 3297-3308. |
[7] | 秦静, 孙法莉, HUI Fang, 汪祖民, 高兵, 季长清. 可穿戴脑电图设备关键技术及其应用综述[J]. 《计算机应用》唯一官方网站, 2022, 42(4): 1029-1035. |
[8] | 彭禹, 宋耀莲, 杨俊. 基于数据增强的运动想象脑电分类[J]. 《计算机应用》唯一官方网站, 2022, 42(11): 3625-3632. |
[9] | 霍首君, 郝琰, 石慧宇, 董艳清, 曹锐. 基于深度卷积网络的运动想象脑电信号模式识别[J]. 计算机应用, 2021, 41(4): 1042-1048. |
[10] | 罗飞, 刘鹏飞, 罗元, 朱思蒙. 多特征融合的运动想象脑电特征提取方法[J]. 《计算机应用》唯一官方网站, 2020, 40(2): 616-620. |
[11] | 王超, 朱明, 王敏. 基于管制员认知负荷和改进蚁群算法的扇区动态通行能力评估[J]. 计算机应用, 2018, 38(1): 277-283. |
[12] | 王若凡, 刘静, 王江, 于海涛, 曹亦宾. 基于功率谱及有限穿越可视图的癫痫脑电信号分析算法[J]. 计算机应用, 2017, 37(1): 175-182. |
[13] | 韩敏, 孙卓然. 基于小波变换和AdaBoost极限学习机的癫痫脑电信号分类[J]. 计算机应用, 2015, 35(9): 2701-2705. |
[14] | 艾玲梅 李营. HHT方法在不同思维作业脑电信号分析中的应用[J]. 计算机应用, 2008, 28(12): 3089-3091. |
阅读次数 | ||||||
全文 |
|
|||||
摘要 |
|
|||||