Structural connectome disruption at baseline predicts 6-months post-stroke outcome

Area of Science:

• Neuroscience
• Medical Imaging
• Biomarkers

Background:

• Stroke significantly disrupts structural brain networks, impacting functional recovery.
• Predicting post-stroke functional outcomes is crucial for personalized rehabilitation strategies.

Purpose of the Study:

• To evaluate the efficacy of quantitative imaging biomarkers of post-stroke connectome disruption in predicting six-month functional outcomes.
• To compare the predictive accuracy of connectome disruption models with traditional lesion volume models.

Main Methods:

• Utilized diffusion-weighted MRI to create lesion masks for 40 ischemic stroke subjects.
• Employed the Network Modification (NeMo) Tool to quantify connectome disruption at whole-brain, regional, and pairwise levels.
• Developed Partial Least Squares Regression models to predict cognitive, mobility, and daily activity outcomes at six months post-stroke.

Main Results:

• Connectome disruption models demonstrated higher accuracy than lesion volume models.
• The regional disconnection model best predicted applied cognitive (R²=0.56) and basic mobility (R²=0.70) outcomes.
• The pairwise disconnection model showed the highest accuracy for predicting daily activity (R²=0.72).

Conclusions:

• Quantitative imaging biomarkers of baseline connectome disruption can reliably predict six-month post-stroke functional outcomes.
• The NeMo Tool, using routine MRI, offers a valuable approach for assessing stroke recovery potential.
• Model accuracy is influenced by the anatomical specificity of disconnection metrics and data dimensionality.