Keywords
Motor Imagery
LSTM
Brain-Computer Interface
Abstract
The present study develops a model for recognizing movement intentions from EEG signals using various RNN configurations, with a
focus on LSTM networks. Experiments demonstrate that, with proper sample allocation and class balance, the model achieves 0.9833 ac-
curacy and an ROC area of 0.9992 when training and test data include the same subjects. The best-performing LSTM model—augmented
with a fully connected layer—was configured with a hidden layer size of 334, learning rate of 3.587 × 10−5, 2 layers, dropout of 0.393,
and sequence length of 86. However, when test subjects were completely excluded from training, the model’s accuracy did not exceed
50%, suggesting significant inter-subject variability or limitations in generalization. This work contributes to advancing Brain-Computer
Interfaces for applications such as prosthetic control and provides insights into the prerequisites for effective EEG signal utilization.
References
N. Boutarfaia et al. “Deep Learning for EEG-Based Motor Imagery Classification: Towards Enhanced Human-
Machine Interaction and Assistive Robotics”. In: CEUR Workshop Proceedings. 2023. URL: https : / /
api.semanticscholar.org/CorpusID:270258728.
A. L. Goldberger et al. “PhysioBank, PhysioToolkit, and PhysioNet: Components of a new research resource
for complex physiologic signals”. In: Circulation 101.23 (2000), e215–e220. DOI: 10.1161/01.CIR.
23.e215. URL: https://physionet.org/content/eegmmidb/1.0.0/.
S. Hochreiter and J. Schmidhuber. “Long Short-Term Memory”. In: Neural Computation 9.8 (1997), pp. 1735–
DOI: 10.1162/neco.1997.9.8.1735. URL: https://doi.org/10.1162/neco.1997.
8.1735.
Yimin Hou et al. “GCNs-Net: A Graph Convolutional Neural Network Approach for Decoding Time-Resolved
EEG Motor Imagery Signals”. In: IEEE Transactions on Neural Networks and Learning Systems 35.6 (June
, pp. 7312–7323. ISSN: 2162-2388. DOI: 10.1109/tnnls.2022.3202569. URL: http://dx.
doi.org/10.1109/TNNLS.2022.3202569.
S. Rajwal and S. Aggarwal. “Convolutional Neural Network-Based EEG Signal Analysis: A Systematic
Review”. In: Archives of Computational Methods in Engineering 30 (2023), pp. 3585–3615. DOI: 10.1007/
s11831-023-09920-1. URL: https://link.springer.com/article/10.1007/s11831-
-09920-1.
K. Roots, Y. Muhammad, and N. Muhammad. “Fusion Convolutional Neural Network for Cross-Subject EEG
Motor Imagery Classification”. In: Computers 9.3 (2020), p. 72. DOI: 10.3390/computers9030072.
URL: https://www.mdpi.com/2073-431X/9/3/72.
Anirban S. Roy and Subir K. Banerjee. “Brain–computer interface advancement in neurosciences”. In: Jour-
nal of the Neurological Sciences 409 (2020), p. 116607. DOI: 10.1016/j.jns.2019.116607. URL:
https://www.sciencedirect.com/science/article/pii/S2214751920300098.
Alex Sherstinsky. “Fundamentals of Recurrent Neural Network (RNN) and Long Short-Term Memory (LSTM)
network”. In: Physica D: Nonlinear Phenomena 404 (2020), p. 132306. ISSN: 0167-2789. DOI: https :
//doi.org/10.1016/j.physd.2019.132306. URL: https://www.sciencedirect.com/
science/article/pii/S0167278919305974.
Yifei Zhang et al. “Cascade and Parallel Convolutional Recurrent Neural Networks on EEG-based Intention
Recognition for Brain Computer Interface”. In: arXiv preprint arXiv:1708.06578 (2021).