Imagined Movement Recognition in People with Disabilities Using Common Sparse Spatio Spectral Pattern (CSSSP) and Sequential Features Selection (SFS)

Subject Areas : Biomedical Signal Processing

Alireza Pirasteh ¹ , Manouchehr Shamseini Ghiyasvand ² , Majid Pouladian ³

1 - Department of Biomedical Engineering, South Tehran Branch, Islamic Azad University, Tehran, Iran
2 - Department of Biomedical Engineering, South Tehran Branch, Islamic Azad University, Tehran, Iran
3 - Department of Biomedical Engineering, Science and Research Branch, Islamic Azad University, Tehran, Iran

Received: 2023-03-07 Accepted : 2023-06-08 Published : 2025-01-05

Keywords: EEG Signal Processing, CSP, CSSP, CSSSP, SFS, Features Extraction, Motor Imagery.,

Abstract :

Motor Imagery is a mental process that includes preparation for movement. The brain interface system intends to prepare direct connectivity between the brain and the computer to be aware of the requests of an individual and use them as a control signal for external devices. Motion imaging events occur in the three main frequency bands: beta, mu, and gamma. After preprocessing the EEG data, the next step is to apply various types of filters in order to reduce any residual noise present in the signal. Numerous functional imaging studies showed that motion-imaging results from the specific activation of neural circuits involved in the early stages of motor control. Studies have shown that the CSP algorithm performs better than other algorithms. Due to the lack of a suitable frequency band, the results of the frequency-dependent CSP method are not satisfactory, so the CSSP is similar to the FIR filter, but since this filter does not have all the coefficients of an FIR filter, the presence of noise in the EEG signal can lead to suboptimal definition of the frequency filter. The CSSSP algorithm was used to solve this problem. With using sequential feature selection for feature extraction, it was revealed that CSSSP performance has been better compared to the CSP and CSSP in most cases and the average accuracy was 92.55%. 

highlight:

We use a new method based on CSP, but with the aim of decrease drawbacks, so the CSSSP method was used.

CSSSP performs both spatial optimization and frequency optimization.

CSSSP simultaneously optimizes a flexible FIR filter with CSP analysis.

The CSSSP method with sequential feature selection (SFS) has shown better performance than CSP and CSSP.

References:

[1] H. Yadav and S. Maini, "Electroencephalogram based brain-computer interface: Applications, challenges, and opportunities," Multimedia Tools and Applications, vol.. 82, no. 30, pp. 1-45, 2023, doi: 10.1007/s11042-023-15653-x.
[2] G. Dornhege, B. Blankertz, M. Krauledat, F. Losch, G. Curio and K. -. Muller, "Combined Optimization of Spatial and Temporal Filters for Improving Brain-Computer Interfacing," in IEEE Transactions on Biomedical Engineering, vol. 53, no. 11, pp. 2274-2281, Nov. 2006, doi: 10.1109/TBME.2006.883649.
[3] J. K. Feng et al., "An Optimized Channel Selection Method Based on Multifrequency CSP-Rank for Motor Imagery-Based BCI System," Computational Intelligence and Neuroscience, vol. 2019, p. 8068357, 2019/05/13 2019,doi: 8068357.
[4] J. Khan et al., “Multiclass EEG motor-imagery classification with sub-band common spatial patterns,” EURASIP Journal on Wireless Communications and Networking, Article number: 174, 2019. doi: 10.1186/s13638-019-1497-y.
[5] N. Korhan, Z. Dokur and T. Olmez, "Motor Imagery Based EEG Classification by Using Common Spatial Patterns and Convolutional Neural Networks," Scientific Meeting on Electrical-Electronics & Biomedical Engineering and Computer Science (EBBT), Istanbul, Turkey, 2019, pp. 1-4, doi: 10.1109/EBBT.2019.8741832.
[6] R. Zhang et al., “Using Brain Network Features to Increase the Classification Accuracy of MI-BCI Inefficiency Subject,” IEEE Access, vol. 7, pp. 74490-74499, 2019, doi: 10.1109/ACCESS.2019.2917327.
[7] Y. Guo, Y. Zhang, Z. Chen, Y. Liu and W. Chen, "EEG classification by filter band component regularized common spatial pattern for motor imagery," Biomedical Signal Processing and Control, vol. 59, p. 101917, 2020, doi: 10.1016/j.bspc.2020.101917.
[8] Y. Jun, Ma. Zhengmin and A. Sh. Tao, "Multi-Time and Multi-Band CSP Motor Imagery EEG Features Classification Algorithm," Applied Sciences , vol. 11, no. 21, p. 10294, 2021, doi: 10.3390/app112110294.
[9] X. Geng, D. Li, H. Chen, P. Yu, H. Yan and M .Yue, “An improved features extraction algorithms of EEG signals based on motor imagery brain-computer interface," Alexandria Engineering Journal, vol. 61, no. 6, pp. 4807-4820, June 2022, doi: 0.1016/j.aej.2021.10.034.
[10] A. Pirasteh, M. Shamseini-Ghiyasvand and M. Pouladian, "Determination of the Type of The Imagined Movement of Organs in People with Mobility Disabilities Using Corrected Common Spatial Patterns," Signal Processing and Renewable Energy, vol. 6, no. 2, pp. 17-19, 2022, Issn: 2588-7327
[11] C. Park, C. Took, D.P. Mandic, "Augmented Complex Common Spatial Patterns for Classification of Noncircular EEG from Motor Imagery Tasks," IEEE Transactions on Neural Systems and Rehabilitation Engineering, vol. 22, no. 1, pp. 1-10, 2014 , doi: 10.1109/TNSRE.2013.2294903.
[12] M. H. Bhatti et al., "Soft Computing-Based EEG Classification by Optimal Feature Selection and Neural Networks," in IEEE Transactions on Industrial Informatics, vol. 15, no. 10, pp. 5747-5754, Oct. 2019, doi: 10.1109/TII.2019.2925624.
[13] B. Blankertz et al., "The BCI competition III: validating alternative approaches to actual BCI problems," in IEEE Transactions on Neural Systems and Rehabilitation Engineering, vol. 14, no. 2, pp. 153-159, June 2006, doi: 10.1109/TNSRE.2006.875642.