Dynamic Information Flow Based on EEG and Diffusion MRI in Stroke: A Proof-of-Principle Study

Area of Science:

• Neuroscience
• Biomedical Engineering
• Computational Neuroscience

Background:

• Stroke recovery involves neural network reorganization, but current neuroimaging has limitations.
• Magnetic Resonance Imaging (MRI) offers high spatial but low temporal resolution for brain activity.
• Electroencephalography (EEG) provides excellent temporal but poor spatial resolution, hindering dynamic brain analysis.

Purpose of the Study:

• To assess a novel multimodal brain imaging technique, Variational Bayesian Multimodal Encephalography (VBMEG).
• To improve EEG's spatial resolution for tracking brain information flow and its changes post-stroke.
• To compare dynamic information flow between stroke survivors and able-bodied individuals.

Main Methods:

• VBMEG integrated EEG with anatomical MRI and diffusion-weighted imaging (DWI) constraints.
• Electrical stimuli were applied to index fingers of stroke patients and healthy controls.
• Active source locations and dynamic information flow were precisely identified and estimated.

Main Results:

• VBMEG precisely identified active brain sources with high Variance Accounted For (VAF > 80%).
• Dynamic information flow estimation achieved high VAF (> 90%) in cross-validation.
• Significant differences in information flow were observed between chronic hemiparetic stroke and able-bodied individuals.

Conclusions:

• VBMEG is feasible for revealing brain information flow changes after stroke.
• This method offers an advanced computational approach to track dynamic neural information flow.
• VBMEG may lead to tools for monitoring cortical network changes and aiding stroke rehabilitation.