Biomarkers of stroke recovery using EEG-based resting-state functional connectivity: a systematic review.

The review included only English peer-reviewed articles from major databases, so relevant work in other languages or grey literature may be missing. The included studies used very different EEG processing, connectivity metrics, frequency bands, sensor or source analyses, and patient characteristics, so findings could not be pooled into a single effect estimate. Many studies were small, lacked healthy control groups, and did not stratify patients by cortical or subcortical lesion location, making it hard to separate genuine recovery effects from methodological or patient differences. Longitudinal studies often compared only before and after rehabilitation within the same group, so changes may reflect spontaneous recovery, training, or both. The supplied text reports 30 included articles in the selection results but refers to 31 included studies in a later synthesis table, so the exact number of studies is unclear.

Do not treat EEG resting-state connectivity as a validated clinical biomarker for stroke recovery. The review found heterogeneous, mostly group-level associations and says these findings are not ready for immediate clinical translation.