Development of a CNN-LSTM Deep Learning Model for Motor Imagery EEG Classification for BCI Applications
Received: 22 December 2024 | Revised: 7 February 2025 and 18 February 2025 | Accepted: 21 February 2025 | Online: 1 April 2025
Corresponding author: Mohd Zuki Yusoff
Abstract
Brain-Computer Interface (BCI) systems offer a groundbreaking method for the human brain to directly communicate with external devices, serving applications, such as assistive technology, smart environments, and healthcare. Motor Imagery (MI) brain signals derived from Electroencephalography (EEG) are commonly utilized in various BCI fields. However, accurately classifying MI-based EEG signals remains a significant challenge, with traditional classification techniques struggling to effectively capture both spatial and temporal features, resulting in suboptimal performance. Therefore, this study introduces a novel hybrid Convolutional Neural Network-Long Short-Term Memory (CNN-LSTM) framework designed for EEG-MI task classification. The model combines adaptive learning with optimal training to significantly improve classification performance using the Berlin BCI Dataset 1 from BCI Competition IV. The proposed CNN-LSTM model achieves a classification accuracy of 98.38% on subject independence evaluation. This research compares subject-dependent and subject-independent evaluation with traditional Machine Learning (ML) methods, such as Support Vector Machines (SVM), Random Forest (RF), and Linear Discriminant Analysis (LDA), as well as Deep Learning (DL) models, such as EEGNet, K-nearest Neighbor (KNN), and Convolutional Neural Network (CNN). Extensive evaluations and cross-validation prove the model's superior performance, thus this work sets a benchmark for real-time MI-EEG classification, offering a scalable solution for practical BCI applications.
Keywords:
signal processing, Brain-Computer Interface (BCI), real-time Motor Imagery (MI), Electroencephalography (EEG), Convolutional Neural Networks (CNNs), Long Short-Term Memory (LSTM), hybrid deep learningDownloads
References
B. Blankertz, G. Dornhege, M. Krauledat, K.-R. Müller, and G. Curio, "The non-invasive Berlin Brain–Computer Interface: Fast acquisition of effective performance in untrained subjects," NeuroImage, vol. 37, no. 2, pp. 539–550, Aug. 2007.
R. Zhang et al., "A New Motor Imagery EEG Classification Method FB-TRCSP+RF Based on CSP and Random Forest," IEEE Access, vol. 6, pp. 44944–44950, 2018.
M. Norizadeh Cherloo, H. Kashefi Amiri, and M. R. Daliri, "Ensemble Regularized Common Spatio-Spectral Pattern (ensemble RCSSP) model for motor imagery-based EEG signal classification," Computers in Biology and Medicine, vol. 135, Aug. 2021, Art. no. 104546.
P. Rithwik, V. K. Benzy, and A. P. Vinod, "High accuracy decoding of motor imagery directions from EEG-based brain computer interface using filter bank spatially regularised common spatial pattern method," Biomedical Signal Processing and Control, vol. 72, Feb. 2022, Art. no. 103241.
A. Apicella, P. Arpaia, M. Frosolone, and N. Moccaldi, "High-wearable EEG-based distraction detection in motor rehabilitation," Scientific Reports, vol. 11, no. 1, Mar. 2021, Art. no. 5297.
M. Dai, D. Zheng, R. Na, S. Wang, and S. Zhang, "EEG Classification of Motor Imagery Using a Novel Deep Learning Framework," Sensors, vol. 19, no. 3, Jan. 2019, Art. no. 551.
Z. Cai, T. Luo, and X. Cao, "Multi-branch spatial-temporal-spectral convolutional neural networks for multi-task motor imagery EEG classification," Biomedical Signal Processing and Control, vol. 93, Jul. 2024, Art. no. 106156.
K. Zhang, N. Robinson, S.-W. Lee, and C. Guan, "Adaptive transfer learning for EEG motor imagery classification with deep Convolutional Neural Network," Neural Networks, vol. 136, pp. 1–10, Apr. 2021.
Y. K. Musallam et al., "Electroencephalography-based motor imagery classification using temporal convolutional network fusion," Biomedical Signal Processing and Control, vol. 69, Aug. 2021, Art. no. 102826.
Y. Han, B. Wang, J. Luo, L. Li, and X. Li, "A classification method for EEG motor imagery signals based on parallel convolutional neural network," Biomedical Signal Processing and Control, vol. 71, Jan. 2022, Art. no. 103190.
J. F. Hwaidi and T. M. Chen, "Classification of Motor Imagery EEG Signals Based on Deep Autoencoder and Convolutional Neural Network Approach," IEEE Access, vol. 10, pp. 48071–48081, 2022.
X. Liu, S. Xiong, X. Wang, T. Liang, H. Wang, and X. Liu, "A compact multi-branch 1D convolutional neural network for EEG-based motor imagery classification," Biomedical Signal Processing and Control, vol. 81, Mar. 2023, Art. no. 104456.
D. Park et al., "Spatio-Temporal Explanation of 3D-EEGNet for Motor Imagery EEG Classification Using Permutation and Saliency," IEEE Transactions on Neural Systems and Rehabilitation Engineering, vol. 31, pp. 4504–4513, 2023.
X. Deng, B. Zhang, N. Yu, K. Liu, and K. Sun, "Advanced TSGL-EEGNet for Motor Imagery EEG-Based Brain-Computer Interfaces," IEEE Access, vol. 9, pp. 25118–25130, 2021.
R. Peng et al., "TIE-EEGNet: Temporal Information Enhanced EEGNet for Seizure Subtype Classification," IEEE Transactions on Neural Systems and Rehabilitation Engineering, vol. 30, pp. 2567–2576, 2022.
Y. R. Tabar and U. Halici, "A novel deep learning approach for classification of EEG motor imagery signals," Journal of Neural Engineering, vol. 14, no. 1, Feb. 2017, Art. no. 016003.
S. A. C. Yohanandan, I. Kiral-Kornek, J. Tang, B. S. Mshford, U. Asif, and S. Harrer, "A Robust Low-Cost EEG Motor Imagery-Based Brain-Computer Interface," in 2018 40th Annual International Conference of the IEEE Engineering in Medicine and Biology Society (EMBC), Jul. 2018, pp. 5089–5092.
N. S. Suhaimi, M. Z. Yusoff, and M. N. M. Saad, "Artificial Neural Network Analysis On Motor Imagery Electroencephalogram," in 2022 IEEE 5th International Symposium in Robotics and Manufacturing Automation (ROMA), Malacca, Malaysia, Aug. 2022, pp. 1–6.
A. M. Roy, "Adaptive transfer learning-based multiscale feature fused deep convolutional neural network for EEG MI multiclassification in brain–computer interface," Engineering Applications of Artificial Intelligence, vol. 116, Nov. 2022, Art. no. 105347.
R. Zhang, Q. Zong, L. Dou, X. Zhao, Y. Tang, and Z. Li, "Hybrid deep neural network using transfer learning for EEG motor imagery decoding," Biomedical Signal Processing and Control, vol. 63, Jan. 2021, Art. no. 102144.
D. Li, J. Wang, J. Xu, X. Fang, and Y. Ji, "Cross-Channel Specific-Mutual Feature Transfer Learning for Motor Imagery EEG Signals Decoding," IEEE Transactions on Neural Networks and Learning Systems, vol. 35, no. 10, pp. 13472–13482, Oct. 2024.
P. Autthasan et al., "MIN2Net: End-to-End Multi-Task Learning for Subject-Independent Motor Imagery EEG Classification," IEEE Transactions on Biomedical Engineering, vol. 69, no. 6, pp. 2105–2118, Jun. 2022.
Y. Hu et al., "A Cross-Space CNN With Customized Characteristics for Motor Imagery EEG Classification," IEEE Transactions on Neural Systems and Rehabilitation Engineering, vol. 31, pp. 1554–1565, 2023.
C. Uyulan, "Development of LSTM&CNN based hybrid deep learning model to classify motor imagery tasks," Communications in Mathematical Biology and Neuroscience, vol. 2021, no. 4, 2021.
Z. Khademi, F. Ebrahimi, and H. M. Kordy, "A review of critical challenges in MI-BCI: From conventional to deep learning methods," Journal of Neuroscience Methods, vol. 383, Jan. 2023, Art. no. 109736.
Z. Khademi, F. Ebrahimi, and H. M. Kordy, "A transfer learning-based CNN and LSTM hybrid deep learning model to classify motor imagery EEG signals," Computers in Biology and Medicine, vol. 143, Apr. 2022, Art. no. 105288.
M. A. Alsuwaiket, "Feature Extraction of EEG Signals for Seizure Detection Using Machine Learning Algorithms," Engineering, Technology & Applied Science Research, vol. 12, no. 5, pp. 9247–9251, Oct. 2022.
H. Li, M. Ding, R. Zhang, and C. Xiu, "Motor imagery EEG classification algorithm based on CNN-LSTM feature fusion network," Biomedical Signal Processing and Control, vol. 72, Feb. 2022, Art. no. 103342.
J. Wang, S. Cheng, J. Tian, and Y. Gao, "A 2D CNN-LSTM hybrid algorithm using time series segments of EEG data for motor imagery classification," Biomedical Signal Processing and Control, vol. 83, May 2023, Art. no. 104627.
Downloads
How to Cite
License
Copyright (c) 2025 Mohd Zuki Yusoff, Aaqib Raza
This work is licensed under a Creative Commons Attribution 4.0 International License.
Authors who publish with this journal agree to the following terms:
- Authors retain the copyright and grant the journal the right of first publication with the work simultaneously licensed under a Creative Commons Attribution License that allows others to share the work with an acknowledgement of the work's authorship and initial publication in this journal.
- Authors are able to enter into separate, additional contractual arrangements for the non-exclusive distribution of the journal's published version of the work (e.g., post it to an institutional repository or publish it in a book), with an acknowledgement of its initial publication in this journal.
- Authors are permitted and encouraged to post their work online (e.g., in institutional repositories or on their website) after its publication in ETASR with an acknowledgement of its initial publication in this journal.
