How does the brain reorganize after a stroke? Discover the new biomarker!

When a stroke blocks a brain region, other healthy areas can "take its place." This phenomenon, called neural compensation, was examined with resting‑state functional MRI (rs‑fMRI). Seventeen patients with acute ischemic stroke underwent three scans during rehabilitation, while 51 healthy volunteers provided a reference connectivity model. Researchers compared the connectivity patterns of the lesion regions with those of every healthy voxel, calculating a "network similarity." They found five healthy areas whose similarity to the lesion network dropped by 29‑54% over time. Interestingly, this reduction coincided with improved daily‑living abilities (ADL), suggesting compensation is strongest early on and fades as rehabilitation succeeds. The finding proposes network similarity as a more sensitive biomarker than traditional motor‑network connectivity measures. In practice, tracking how healthy areas diverge from the lesion network could help clinicians gauge rehab progress and tailor interventions.

Neural compensation has traditionally been inferred from increased functional connectivity in non‑lesioned regions, yet without a direct link to behavioral recovery. In this work, Lesion Network Mapping (LNM) was combined with rs‑fMRI to generate voxel‑wise lesion connectivity maps. Network similarity was quantified by Pearson correlation between the lesion's connectivity vector and that of each healthy voxel, yielding a similarity distribution per patient. Longitudinal analysis revealed a significant decrease in similarity in five cortical areas (likely involved in sensorimotor and attention networks), with effect sizes η² between 0.10 and 0.20. The strong statistical association with ADL scores (p < 0.001, η² = 0.90) indicates that similarity reduction reflects a shift from active compensation to functional consolidation. This approach overcomes the limits of absolute connectivity measures by considering the lesion's topological context against a healthy reference atlas, providing a more sensitive biomarker for tracking cerebral plasticity during rehabilitation. Moreover, the method is scalable to multicenter studies and could be combined with targeted neuromodulation, such as transcranial magnetic stimulation, to boost emerging compensatory regions.