Brain Connectivity Measures Improve Modeling of Functional Outcome After Acute Ischemic Stroke

Area of Science:

• Neuroscience
• Medical Imaging
• Network Science

Background:

• Predicting long-term functional outcomes after acute ischemic stroke remains a clinical challenge.
• Connectomics, representing brain connectivity as a graph, offers potential for improved prediction models.
• Rich club organization, a critical brain network backbone, may be impacted by stroke lesions.

Purpose of the Study:

• To assess the impact of acute ischemic stroke lesions on the brain's rich club organization and functional connectomes.
• To determine the relationship between network topological properties (characteristic path length, L) and post-stroke functional outcomes.
• To evaluate the added value of network topology metrics in predicting stroke outcomes compared to traditional markers.

Main Methods:

• Analysis of functional connectomes derived from 3 anatomic atlases in 41 acute ischemic stroke patients.
• Calculation of characteristic path length (L) for each connectome.
• Manual determination of affected rich club regions from diffusion-weighted images and assessment of L and region counts in outcome prediction models (modified Rankin Scale, NIHSS).

Main Results:

• Lower characteristic path length (L) was generally associated with better functional outcomes.
• Inclusion of the number of affected rich club regions significantly increased the explained variance (R^2) in outcome prediction models.
• Model R^2 increased 1.3- to 2.6-fold for NIHSS and modified Rankin Scale scores when network topology metrics were included alongside age and lesion volume.

Conclusions:

• Network topology information, specifically related to the rich club, can enhance the modeling of post-stroke functional outcomes.
• This proof-of-concept study demonstrates the utility of connectomics in stroke outcome prediction.
• Larger prospective studies are needed to validate this approach for clinical application in stroke outcome prediction.