Mu and Beta Oscillatory Changes during a motor task following Rehabilitation in Chronic MCA Stroke: Insights from EEG

Dataset Summary This dataset includes electroencephalographic (EEG) recordings and clinical data from a cohort of chronic ischaemic left AMC stroke patients and healthy controls. A total of 10 stroke patients (6 males, 4 females; mean age 55.9 ± 14.2 years) were initially recruited from an ongoing study. EEG measurements were collected before and after a rehabilitation intervention. One patient was lost to follow-up. Additionally, 7 right-handed healthy controls (1 male, 6 females; mean age 46.3 ± 23.6 years) with no neurological history were included for comparison. EEG data were acquired using a 64-channel system (10–20 international system) at a sampling rate of 1000 Hz. Recordings were conducted under resting-state conditions (eyes closed, 5 minutes) and during motor tasks, including motor imagery and active hand movements. The experimental motor paradigm consisted of 140 trials per subject, combining visual and auditory cues to guide movement execution or imagination. Preprocessing was performed using EEGLAB, ERPLAB, and MATLAB scripts. Steps included visual inspection, bad channel removal and interpolation, notch filtering (50 Hz), band-pass filtering (0.5–100 Hz), artifact rejection, and Independent Component Analysis (ICA). Data were segmented into epochs for both resting-state and task-related analyses. Quantitative EEG (qEEG) measures were computed using spectral power density (Welch method) across standard frequency bands (delta, theta, alpha, beta). Derived indices include relative delta power, delta/alpha ratio (DAR), delta+theta/alpha+beta ratio (DTABR), and Brain Symmetry Index (BSI). Event-related desynchronization/synchronization (ERD/ERS) and event-related potentials (ERP) were analyzed using defined time windows for baseline, motor imagery, and motor execution. Statistical analyses employed non-parametric tests (Mann–Whitney U, Wilcoxon) due to the small sample size, with a significance threshold of p < 0.05. Analyses were exploratory and focused on identifying EEG markers differentiating stroke patients and healthy controls, as well as pre/post intervention changes. A complete guide of "how to" filtrate the EEG using MATMAB is included in a word file (in Spanish) with images to guide you through the process. The matlab archives with the code are included in the zip folder.