While the multidimensional features of electroencephalographic (EEG) signals have proven to be a valuable source of information, the development of a comprehensive diagnostic tool remains elusive due to variability of responses as observed within the subjects and epochs. We investigate whether ordinal-pattern-based complexity measures of EEG signals can capture stroke-related alterations in motor imagery (MI) tasks. EEG recordings from 36 stroke patients (acute and minor) and 36 healthy controls were analyzed using permutation entropy (PE), a robust symbolic measure of temporal irregularity. Stroke patients perform left- and right-hand MI tasks, while controls are recorded only under eye-open MI and eye-closed resting conditions. Results show that resting-state EEG from healthy participants exhibits low PE values, reflecting structured and regular dynamics, whereas eye-open MI EEG from the same cohort produces high PE values consistent with near-maximally complex, information-rich neural dynamics. Stroke patients demonstrate intermediate PE values during MI tasks, suggesting altered but partially preserved physiological complexity. These findings indicate that entropy-based measures can distinguish between healthy and stroke-related neural dynamics, providing potential biomarkers for tailoring brain-computer interface (BCI) driven rehabilitation strategies.
Bose et al. (Mon,) studied this question.