Reshaped functional connectivity gradients in acute ischemic stroke

Ischemic stroke disrupts cerebral blood flow, triggering both structural and functional brain alterations that often lead to behavioral impairments. Although the primary damage typically affects specific regions, the resulting changes in brain organization extend widely across the cortex and are high-dimensional in nature. The underlying mechanisms driving these widespread functional disturbances and their connection to behavioral symptoms remain largely underexplored.

Functional connectivity gradients offer a reproducible and robust way to represent brain organization in a low-dimensional space, simplifying complex functional variation into a few interpretable axes. In this study, we examined how stroke affects this canonical gradient space by aligning individual patients’ connectivity profiles to a normative template derived from healthy controls. We then measured how far each region deviated from its expected position, using these deviations to assess their relevance for behavioral outcomes. Crucially, we accounted for stroke-induced hemodynamic delays to better understand their influence.

Our analysis showed that correcting for hemodynamic lags significantly improved gradient accuracy, particularly in the second gradient, which encompasses visual and somatomotor functions. Notable functional deviations were observed within the somatomotor, visual, and ventral attention networks—areas linked to behavioral impairments following stroke.

We also investigated hemispheric asymmetries in these deviations. Interestingly, unaffected hemispheres retained typical asymmetry patterns, whereas stroke-affected hemispheres showed marked disruptions. Moreover, right-hemisphere lesions resulted in more localized functional alterations compared to left-sided damage.

Overall, our findings highlight two key insights: (1) adjusting for hemodynamic delays increases the explanatory power of connectivity gradients, and (2) behavioral deficits and altered hemispheric asymmetries can be traced to shifts in region-specific connectivity patterns within a low-dimensional, interpretable framework. This supports the idea that post-stroke brain reorganization follows predictable trajectories along fundamental axes of brain function—and that these shifts are not entirely independent of underlying white matter damage.

Authors: Cemal Koba, Joan Falcó-Roget, Alessandro Crimi

DOI: 10.1016/j.nicl.2025.103755

Keywords: Stroke, fMRI, Gradients, Temporal lag