PHASE SPACE RECONSTRUCTION FOR BRAIN STATE CLASSIFICATION BY EEG SIGNALS
DOI:
https://doi.org/10.34185/1991-7848.itmm.2025.01.060Keywords:
EEG, phase space reconstruction, recurrence analysis, chaotic dynamics, brain state classification.Abstract
This thesis explores the classification of brain states based on EEG data, focusing on the distinction between relaxation and concentration. A classification approach using recurrence plot analysis, a method from chaos theory, is compared with traditional spectral analysis. The optimal phase space reconstruction parameters were determined: a delay equal 25 ms and an dimension of embedding space equal 4. These values align with spectral characteristics of the EEG signal, confirming their physiological relevance. The study suggests that these parameters can be used to develop differential equations describing chaotic brain activity. The findings are relevant for EEG analysis in portable devices, brain-computer interfaces, and cognitive training applications.
References
V. Ye. Belozyorov. On novel conditions of chaotic attractors existence in autonomous polynomial dynamical systems. Nonlinear Dynamics. Вип. 91, № 4. С. 2435–2452. DOI:10.1007/s11071-017-4023-y.
EEG Motor Movement/Imagery Dataset. URL: https://archive.physionet.org/pn4/eegmmidb/ (accessed 02/21/2024).
Myers A., Khasawneh F. A. On the automatic parameter selection for permutation entropy. Chaos: An Interdisciplinary Journal of Nonlinear Science. Vol. 30, Issue 3. P. 033130. DOI:10.1063/1.5111719.
