Decoding Multi-Class EEG Signals of Hand Movement Using Multivariate Empirical Mode Decomposition and Convolutional Neural Network

Abstract

Brain-computer interface (BCI) is a technology that connects the human brain and external devices. Many studies have shown the possibility of using it to restore motor control in stroke patients. One specific challenge of such BCI is that the classification accuracy is not high enough for multi-class movements. In this study, by using Multivariate Empirical Mode Decomposition (MEMD) and Convolutional Neural Network (CNN), a novel algorithm (MECN) was proposed to decode EEG signals for four kinds of hand movements. Firstly, the MEMD was used to decompose the movement-related electroencephalogram (EEG) signals to obtain the multivariate intrinsic empirical functions (MIMFs). Then, the optimal MIMFs fusion was performed based on sequential forward selection algorithm. Finally, the selected MIMFs were input to the CNN model for discriminating four kinds of hand movements. The average classification accuracy of thirteen subjects over the six-fold cross-validation reached 81.14% for 2s-data before the movement onset and 81.08% for 2s-data after

Cortical signals analysis to recognize intralimb mobility using modified RNN and various EEG quantities

“…Previous research has yielded promising results in identifying two-class hand movements with one hand, with a classification accuracy of approximately 80 %. However, the accuracy of multi-class hand motions of one hand ranged between 50 % and 70 %, which was insufficient to justify the use of rehabilitation training [ 38 ]. R. Ma et al [ 39 ], developed an EEG-BCI paradigm that matched the steps of both one-sided lower and opposite-sided upper limb motions (also known as compound-limb movements).…”

“… Refer Year Classification scheme Classification Performance Dillen et al [ 23 ] 2022 Within Subject scheme for Inter limb movement. 0.844 ± 0.088 G. Zhang et al [ 26 ] 2019 Within Subject scheme Across subjects 98.3 ± 0.9 % 83.2 ± 1.2 % Liu et al [ 20 ] 2018 Within Subject scheme 84.44 ± 14.56 % Tao et al [ 38 ] 2022 Within Subject scheme 81.14 ± 6.76 % Chaisaen et al [ 24 ] 2020 Within Subject scheme 82.73 ± 2.54 % J. Hoon te al [ 25 ] 2020 Within Subject scheme 86 ± 9.0 % R. Ma [ 39 ] 2023, Within Subject scheme Unilateral Lower limb 89.02 ± 12.84 % 62.68 ± 4.54 % Yaqi Chu et al [ 40 ] 2023 Within Subject scheme 80.50 % L. Gu et al [ 41 ] 2023 Within Subject scheme 62.94 % …”