Voronezh, Russian Federation
Voronezh, Russian Federation
UDC 004.032.26
A compact hardware and software system based on a neuroheadset with dry electrodes and wireless data transmission developed. The system designed to generate control inputs from the operator while simultaneously monitoring their functional state in real time and integrated into complex technological systems. The results of a study on the effectiveness of binary classification of operator motor patterns using several classifiers operating in combination with different optimization algorithms presented. The following classifiers analyzed Rosenblatt perceptron, linear discriminant analysis, and convolutional neural network. A classifier architecture based on a ResNet-type convolutional neural network consisting of eighteen repeating macrolayers is proposed. Using the Accuracy, Precision, Recall, and F1-score metrics, we analyzed the impact of various optimization algorithms (adaptive moment estimation, Levenberg-Marquardt with the proposed upgrade, stochastic gradient descent, and Broyden-Fletcher-Goldfarb-Shanno) on classification results. The best online performance demonstrated by a combination of a convolutional neural network-based classifier and the adaptive moment estimation algorithm. The classification success rate using the Accuracy metric was approximately 66%. The obtained results found to exceed typical results for mobile handheld brain-computer interfaces operating in real-time (online).
asynchronous brain-computer interface, motor images, binary classification, Rosenblatt perceptron, linear discriminant analysis, convolutional neural network
1. Suryotrisongko H., Samopa F. Evaluating openbci spiderclaw v1 headwear's electrodes placements for brain-computer interface (BCI) motor imagery application // Procedia Computer Science. 2015. V. 72. P. 398-405. https://www.doi.org/10.1016/j.procs.2015.12.155
2. Kapralov N. V., Nagornova Z. V., Shemyakina N. V. Classification methods for EEG patterns of imaginary movements // Informatics and Automation. 2021. V. 20(1). P. 94-132. (In Russ.).
3. Golubinskii A. N., Tolstikh A. A. O primenenii svertochnikh neironnikh setei dlya klassifikatsii motoriki na osnove signalov EEG interfeisa mozg-kompyuter // Naukosfera. 2021. T. 2-1. S. 85-88.
4. Pavlenko D. V., Tataris Sh. E., Ovcharenko V. V. Primenenie glubokogo obucheniya v interfeisakh mozg–kompyuter dlya raspoznavaniya dvizhenii // Programmnie produkti i sistemi. 2024. T. 37(2). S. 164-169.
5. Echtioui A., Mlaouah A., Zouch W., Ghorbel M., Mhiri C., Hamam H. A novel convolutional neural network classification approach of motor-imagery EEG recording based on deep learning. Applied Sciences. 2021; 11(21): 9948. https://www.doi.org/10.3390/app11219948
6. Zhuravlev D.V., Golubinsky A.N., Tolstykh A.A. Development of a methodology for setting parameters of brain-computer interfaces for conducting experiments on classification of motor images in the OpenVibe program // Biomedical radio electronics. 2025. Vol. 28(3). pp. 15-30.Chi J., Needell D. Linear Discriminant Analysis with the Randomized Kaczmarz Method // SIAM Journal on Matrix Analysis and Applications. 2025. V. 46. P. 94-120. https://www.doi.org/10.1137/23M155493X
7. Chi J., Needell D. Linear Discriminant Analysis with the Randomized Kaczmarz Method // SIAM Journal on Matrix Analysis and Applications. 2025. V. 46. P. 94-120. https://www.doi.org/10.1137/23M155493X
8. Ji L., Wei Z., Hao J., Wang C. An intelligent diagnostic method of ECG signal based on Markov transition field and a ResNet // Computer Methods and Programs in Biomedicine. 2023. V. 242(7). P. 107784. https://www.doi.org/10.1016/j.cmpb.2023.107784
9. Nawi N. M., Ransing M. R., Ransing R. S. An improved learning algorithm based on the Broyden-Fletcher-Goldfarb-Shanno (BFGS) method for back propagation neural networks // Sixth International Conference on Intelligent Systems Design and Applications (IEEE). 2006. P. 152-157. https://www.doi.org/10.1109/ISDA.2006.95
10. Miconi T. Hebbian learning with gradients: Hebbian convolutional neural networks with modern deep learning frameworks. arXiv. URL: https://arxiv.org/abs/2107.01729 [Accessed 21st January 2026].
11. Rubio J. de J. Stability Analysis of the Modified Levenberg-Marquardt Algorithm for the Artificial Neural Network Training // IEEE Transactions on Neural Networks and Learning Systems. 2021. V. 32(8). P. 3510-3524. https://www.doi.org/10.1109/TNNLS.2020.3015200
12. Panteleev A.V., Lobanov A.V. Gradient optimization methods in machine learning for dynamic system parameter identification // Data modeling and analysis. 2019. V. 9(4). P. 88-99. https://doi.org/10.17759/mda.2019090407
13. Ruder S. An Overview of Gradient Descent Optimization Algorithms arXiv:1609.04747v2 [cs.LG] 15 Jun 2017 [Accessed 21st January 2026].
14. Floudas C.A., Pardalos P.M., Adjimann C.S., Esposito W.R., Gumus Z.H., Harding S.T., Schweiger C.A. Handbook of test problems in local and global optimization. 1999. V. 67. Springer US. 442 p. https://titan.princeton.edu/TestProblems/
15. Tjoa I.–B., Biegler L.T. Simultaneous solution and optimization strategies for parameter estimation of differential–algebraic equation systems // Industrial & Engineering Chemistry Research. 1991. V. 30(2). P. 376-385. https://doi.org/10.1021/ie00050a015
