Mapping Brain Function After Stroke: From Early Disorganization to Late Compensation

Abstract

Introduction

Electroencephalography (EEG) is useful to capture changes in brain function after stroke. How these relate to time after stroke and deficit severity remains unclear. We examined these issues by testing specific hypotheses using cortical power and functional connectivity.

Methods

This cross-sectional study included 100 stroke patients and 45 healthy controls. Cortical power and connectivity (coherence) were extracted from a 3-minute resting-state dense-array (256-lead) EEG. Analysis included subgroups based on time post-stroke (<or≥ 90 days) and motor status (Fugl-Meyer score ≤or> 40) using robust permutation tests and partial Spearman’s correlations.

Results

Early after stroke, power was supranormal in low frequency bands and subnormal in high frequency bands, while brain connectivity was widely subnormal. With greater time post-stroke, beta and gamma connectivity were higher bilaterally, and brain connectivity was supranormal in the late phase across all frequency bands. Motor status was related to cortical power only in cortico–subcortical patients and was related to connectivity in a manner that varied according to severity of motor deficit and lesion location.

Conclusions

Results support use of EEG measures as biomarkers of brain reorganization after stroke and emphasize the need for a nuanced understanding according to individual details. Current findings indicate that motor deficits are best understood using a network connectivity approach and highlight the need to interpret EEG measures post-stroke within the specific context (time post-stroke, deficit severity, and lesion location) in which they are acquired to maximize their utility as biomarkers to guide personalized rehabilitation.