A connectome-based functional magnetic resonance imaging study of visuospatial analogical reasoning in stroke.

The study is cross-sectional and correlational, so it cannot prove that high connectivity causes poor reasoning or that the stroke caused the connectivity changes. Participants were limited to those with relatively small lesions, because large lesions often made fMRI preprocessing fail. This means the findings may not apply to patients with severe, large strokes. The regions with high connectivity were not directly damaged, so the mechanism is likely indirect or compensatory, but the study cannot determine if this hyperconnectivity is helpful or harmful. The sample size was small (58 patients), and the clustering analysis was data-driven, which can be unstable. All patients received standard physiotherapy, occupational therapy, and speech therapy as needed, so the specific contribution of any single therapy to the observed cognitive state cannot be separated.

The authors declared no conflicts of interest. The study was funded by JSPS KAKENHI grants.

Do not assume that high functional connectivity in the parietal network indicates a healthy or intact system. In this study, higher connectivity was associated with worse performance and longer solving times, suggesting it may represent inefficient compensatory processing rather than preserved function.