Related Experiment Video
Updated: Aug 5, 2026
06:37
Electroencephalography Network Indices as Biomarkers of Upper Limb Impairment in Chronic Stroke
Published on: July 14, 2023
EMBC Special Issue: Structural Path Lengths Predict Post-Stroke Upper Limb Functional Outcome: Activation Likelihood
IEEE Transactions on Bio-Medical Engineering
|July 31, 2026
Summary
Baseline clinical and neurophysiological measures are the strongest predictors of post-stroke motor recovery. Task-specific brain atlases offer limited additional value over established clinical predictors for forecasting upper-limb function after stroke.
Area of Science:
- Neuroscience
- Neurology
- Medical Imaging
Background:
- Stroke recovery of upper-limb motor function is a critical clinical challenge.
- Predictive modeling for post-stroke motor outcomes aids in treatment planning and prognosis.
- Brain atlases and lesion network mapping are emerging tools in understanding stroke recovery.
Purpose of the Study:
- To compare the predictive utility of shortest structural path length (SSPL) features derived from whole-brain versus task-specific atlases for forecasting post-stroke upper-limb motor outcomes.
- To evaluate the incremental prognostic value of SSPL features beyond established clinical and neurophysiological measures.
Main Methods:
- Retrospective analysis of 142 stroke patients with baseline clinical (NIHSS, SAFE, FMUE), neurophysiological (MEP status), and demographic data.
- Quantification of SSPL features using the Lesion Quantification Toolkit with HCP-842 atlas connectivity.
- Extraction of SSPL features from whole-brain (Schaefer 100) and task-specific (SMAA) atlases.
- Training regression models to predict 12-week upper-limb outcomes and assessing feature importance using SHAP.
Main Results:
- Baseline clinical models demonstrated superior predictive performance compared to SSPL-based models.
- SSPL features from the task-specific SMAA atlas showed comparable performance to whole-brain atlas features.
- Neither whole-brain nor SMAA-derived SSPL features provided significant additional predictive value when combined with baseline clinical measures.
Conclusions:
- Established clinical and neurophysiological measures (MEP, motor impairment scales) are the primary predictors of upper-limb motor recovery post-stroke.
- While task-specific parcellations offer a focused view of motor network disruption, their prognostic utility is limited beyond existing clinical predictors.
