A Novel Simplified Convolutional Neural Network Classification Algorithm of Motor Imagery EEG Signals Based on Deep Learning
Source abstract
Abstract
Left and right hand motor imagery electroencephalogram (MI-EEG) signals are widely used in brain-computer interface (BCI) systems to identify a participant intent in controlling external devices. However, due to a series of reasons, including low signal-to-noise ratios, there are great challenges for efficient motor imagery classification. The recognition of left and right hand MI-EEG signals is vital for the application of BCI systems. Recently, the method of deep learning has been successfully applied in pattern recognition and other fields. However, there are few effective deep learning algorithms applied to BCI systems, particularly for MI based BCI. In this paper, we propose an algorithm that combines continuous wavelet transform (CWT) and a simplified convolutional neural network (SCNN) to improve the recognition rate of MI-EEG signals. Using the CWT, the MI-EEG signals are mapped to time-frequency image signals. Then the image signals are input into the SCNN to extract the features and classify them. Tested by the BCI Competition IV Dataset 2b, the experimental results show that the average classification accuracy of the nine subjects is 83.2%, and the mean kappa value is 0.651, which is 11.9% higher than that of the champion in the BCI Competition IV. Compared with other algorithms, the proposed CWT-SCNN algorithm has a better classification performance and a shorter training time. Therefore, this algorithm could enhance the classification performance of MI based BCI and be applied in real-time BCI systems for use by disabled people.
Source: crossref · Quality: baseline
Research context
OpenAlex topics: EEG and Brain-Computer Interfaces · Gaze Tracking and Assistive Technology · Neuroscience and Neural Engineering
SciVX discovery tags are terminology-based. OpenAlex topics are source-provided classifications.
Affiliations
School of Computer and Communication Engineering, Changsha University of Science and Technology, Changsha 410114, China
Hunan Provincial Key Laboratory of Intelligent Processing of Big Data on Transportation, Changsha University of Science and Technology, Changsha 410114, China
School of Software, South China Normal University, Guangzhou 510631, China
Funding
National Natural Science Foundation of China · Natural Science Foundation of Hunan Province · Scientific 249 Research Fund of Hunan Provincial Education Department · "Double First-class" International Cooperation and Development Scientific Research Project of Changsha University of Science and Technology · Young Teacher Growth Plan Project of Changsha University of Science and Technology
References (50)
Show source-provided references
1. Brain–computer interfaces for communication and control10.1016/S1388-2457(02)00057-3 ↗
2. Brain–computer interfaces: Communication and restoration of movement in paralysis10.1113/jphysiol.2006.125633 ↗
3. EEG-based BCI and video games: A progress report
4. Improving performance of Devanagari script input-based P300 speller using deep learning10.1109/TBME.2018.2875024 ↗
5. Multi-ganglion ANN based feature learning with application to P300-BCI signal classification10.1016/j.bspc.2014.12.007 ↗
6. A wearable SSVEP-based BCI system for quadcopter control using head-mounted device10.1109/ACCESS.2018.2825378 ↗
7. Kwak, N.S., Müller, K.R., and Lee, S.W. (2017). A convolutional neural network for steady state visual evoked potential classification under ambulatory environment. PLoS ONE, 12.10.1371/journal.pone.0172578 ↗
8. A deep learning scheme for motor imagery classification based on restricted boltzmann machines10.1109/TNSRE.2016.2601240 ↗
9. The extraction of motion-onset VEP BCI features based on deep learning and compressed sensing10.1016/j.jneumeth.2016.11.002 ↗
10. Machine learning for real-time single-trial EEG-analysis: From brain–computer interfacing to mental state monitoring10.1016/j.jneumeth.2007.09.022 ↗
11. Non-linear analysis of EEG signals at various sleep stages10.1016/j.cmpb.2005.06.011 ↗
12. Motor imagery EEG recognition based on conditional optimization empirical mode decomposition and multi-scale convolutional neural network10.1016/j.eswa.2020.113285 ↗
13. Wu, S.L., Wu, C.W., Pal, N.R., Chen, C., Chen, S., and Lin, C. (2013, January 16–19). Common spatial pattern and linear discriminant analysis for motor imagery classification. Proceedings of the 2013 IEEE Symposium on Computational Intelligence, Cognitive Algorithms, Mind, and Brain, Singapore.
14. Structured optimal graph based sparse feature extraction for semi-supervised learning10.1016/j.sigpro.2020.107456 ↗
15. An Automatic Channel Selection Approach for ICA-Based Motor Imagery Brain Computer Interface10.1007/s10916-018-1106-3 ↗
16. Sakhavi, S., Guan, C., and Yan, S. (2015, January 1–4). Parallel convolutional-linear neural network for motor imagery classification. Proceedings of the 2015 23rd European Signal Processing Conference (EUSIPCO), Nice, France.10.1109/EUSIPCO.2015.7362882 ↗
17. An end-to-end deep learning approach to MI-EEG signal classification for BCIs10.1016/j.eswa.2018.08.031 ↗
18. Interpretable deep neural networks for single-trial EEG classification10.1016/j.jneumeth.2016.10.008 ↗
19. A novel hybrid deep learning scheme for four-class motor imagery classification10.1088/1741-2552/ab3471 ↗
20. Li, M.A., Zhang, M., and Sun, Y.J. (2016, January 30–31). A novel motor imagery EEG recognition method based on deep learning. Proceedings of the 2016 International Forum on Management, Education and Information Technology Application, Guangzhou, China.10.2991/ifmeita-16.2016.133 ↗
21. Single-trial EEG classification of motor imagery using deep convolutional neural networks10.1016/j.ijleo.2016.10.117 ↗
22. A novel deep learning approach for classification of EEG motor imagery signals10.1088/1741-2560/14/1/016003 ↗
23. Analysis of EEG records in an epileptic patient using wavelet transform10.1016/S0165-0270(02)00340-0 ↗
24. Deep learning10.1038/nature14539 ↗
25. A convolutional neural network-based linguistic steganalysis for synonym substitution steganography10.3934/mbe.2020055 ↗
26. Strubell, E., Ganesh, A., and McCallum, A. (2019). Energy and policy considerations for deep learning in NLP. arXiv.10.18653/v1/P19-1355 ↗
27. A multi-agent system for the classification of gender and age from images10.1016/j.cviu.2018.01.012 ↗
28. Detecting seam carved images using uniform local binary patterns
29. Dai, M., Zheng, D., Na, R., Wang, S., and Zhang, S. (2019). EEG classification of motor imagery using a novel deep learning framework. Sensors, 19.10.3390/s19030551 ↗
30. Deep learning in spiking neural networks10.1016/j.neunet.2018.12.002 ↗
31. On the information bottleneck theory of deep learning10.1088/1742-5468/ab3985 ↗
32. Deep learning based on Batch Normalization for P300 signal detection10.1016/j.neucom.2017.08.039 ↗
33. Maddula, R., Stivers, J., Mousavi, M., and Ravindran, S. (2017, January 18–22). Deep Recurrent Convolutional Neural Networks for Classifying P300 BCI signals. Proceedings of the GBCIC, Graz, Austria.
34. Aznan, N.K.N., Bonner, S., Connolly, J., Moubayed, N., and Breckon, T. (2018, January 7–10). On the classification of SSVEP-based dry-EEG signals via convolutional neural networks. Proceedings of the 2018 IEEE International Conference on Systems, Man, and Cybernetics, Miyazaki, Japan.
35. Kumar, S., Sharma, A., Mamun, K., and Tsunoda, T. (2016, January 5–6). A deep learning approach for motor imagery EEG signal classification. Proceedings of the 2016 3rd Asia-Pacific World Congress on Computer Science and Engineering, Nadi, Fiji.10.1109/APWC-on-CSE.2016.017 ↗
36. Tayeb, Z., Fedjaev, J., Ghaboosi, N., Richter, C., and Everding, L. (2019). Validating deep neural networks for online decoding of motor imagery movements from EEG signals. Sensors, 19.10.3390/s19010210 ↗
37. Li, M., Zhu, W., Zhang, M., Sun, Y., and Wang, Z. The novel recognition method with optimal wavelet packet and LSTM based recurrent neural networkIn Proceedings of the IEEE International Conference on Mechatronics and Automation, Ningbo, China, 19–21 November 2017.10.1109/ICMA.2017.8015882 ↗
38. EEG-based discrimination between imagination of right and left hand movement10.1016/S0013-4694(97)00080-1 ↗
39. Stockwell-common spatial pattern technique for motor imagery-based Brain Computer Interface design10.1016/j.compeleceng.2018.07.056 ↗
40. Optimized motor imagery paradigm based on imagining Chinese characters writing movement10.1109/TNSRE.2017.2655542 ↗
41. Handwritten Digit Recognition with a Back-Propagation Network
42. Hanin, B. (2019). Universal function approximation by deep neural nets with bounded width and relu activations. Mathematics, 7.10.3390/math7100992 ↗
43. Top-down neural attention by excitation backprop10.1007/s11263-017-1059-x ↗
44. Sun, S., and Zhou, J. (2014, January 6–11). A review of adaptive feature extraction and classification methods for EEG-based brain-computer interfaces. Proceedings of the 2014 International Joint Conference on Neural Networks (IJCNN), Beijing, China.10.1109/IJCNN.2014.6889525 ↗
45. An, X., Kuang, D., Guo, X., and Zhao, Y. (2014, January 3–6). A deep learning method for classification of EEG data based on motor imagery. Proceedings of the International Conference on Intelligent Computing, Taiyuan, China.10.1007/978-3-319-09330-7_25 ↗
46. Optimizing Spatial filters for Robust EEG Single-Trial Analysis10.1109/MSP.2008.4408441 ↗
47. Neurophysiological substrates of stroke patients with motor imagery-based brain-computer interface training10.3109/00207454.2013.850082 ↗
48. Daily training with realistic visual feedback improves reproducibility of event-related desynchronisation following hand motor imagery10.1016/j.clinph.2013.03.006 ↗
49. EEG-based motor imagery analysis using weighted wavelet transform features10.1016/j.jneumeth.2008.09.014 ↗
50. Ma, L., Stückler, J., Wu, T., and Cremers, D. (2018). Detailed Dense Inference with Convolutional Neural Networks via Discrete Wavelet Transform. arXiv.