Topographical measures of functional connectivity as biomarkers for post-stroke motor recovery.
Gavin R Philips, Janis J Daly, José C Príncipe
PMID 28683745WHAT IT FOUND
In chronic stroke survivors, EEG beta-band network measures were associated with upper-limb recovery only after a selected threshold.
The study is exploratory, not tested for clinical use.
Key findings
01With a threshold of one standard deviation above the median, no graph measure change showed a significant relationship with FMUE change in either the full group or the cortically lesioned subset.
02With the t = 0.05 sparsity threshold, the 12.5-25 Hz unaffected-hemisphere intradensity correlated negatively with FMUE change in all subjects (r_s = -0.46; p = 0.015), and 12.5-25 Hz global efficiency, local efficiency, and unaffected-hemisphere intradensity correlated negatively with FMUE change in cortically lesioned subjects (r_s = -0.60; p = 0.017; r_s = -0.66; p = 0.004; r_s = -0.75; p = 0.003).
03Higher initial values of 12.5-25 Hz local efficiency predicted greater FMUE improvement in all subjects (R2 = 0.16; p = 0.030), and higher initial values of 12.5-25 Hz global and local efficiency predicted greater improvement in cortically lesioned subjects (R2 = 0.24; p = 0.047; R2 = 0.25; p = 0.042).
STILL TO COME
How it was doneWhat they foundWhat it means for PTs
Read the rest of this summary
You get three full summaries a month, free, and we do not ask for a card. Search, the TL;DRs and your library stay unlimited either way.
What it does not show
The study included only 30 subjects, and many significant correlations involved a subset of 17 subjects with cortical lesions, so they may not apply to all stroke survivors. All subjects received the same comprehensive intensive intervention, so the study cannot show which treatment component caused recovery or whether biomarker changes were due to one part of the package. The significance of the results depended heavily on the graph threshold: a common threshold showed no significant relationship with FMUE change, while a t = 0.05 threshold showed several significant correlations. Node-specific local efficiency findings were not significant after correction for multiple comparisons and should be treated as exploratory. EEG volume conduction was not fully removed; the authors relied on pre/post contrasts and sparsity, which assumes spurious effects are identical across conditions. Prognostic analysis used baseline biomarker values rather than condition contrasts, which the authors say reduces the effectiveness of their approach to volume conduction. Only two time points were measured, before and after 12 weeks, so the study cannot show how biomarkers change across recovery phases. The subjects were heterogeneous in lesion location and severity, and some had right-arm or non-dominant affected-arm tasks; flipping right-hemisphere data may introduce confounding.
Declared interests
The work was funded by the University of Florida, the U.S. Department of Veterans Affairs, and the National Institute of Neurological Disorders and Stroke. The supplied text does not report author conflicts of interest.
The easy way to misread this
Do not use these EEG graph measures to guide stroke therapy or prognosis. The significant associations appeared only after selecting a t = 0.05 sparsity threshold, a common threshold produced no significant relationships, node-level results were not significant after multiple-comparison correction, and the study did not test clinical use.