Abstract TP257: Advanced Diffusion Trajectories of Lesioned and Contralesional White Matter in Cerebellar vs. Non-Cerebellar Stroke Recovery

Introduction: Cerebellar stroke (CS) mechanisms of cognitive dysfunction remain unclear. Conventional MRI often misses subtle microstructural changes. Advanced diffusion imaging enables longitudinal assessment of normal-appearing tissue, with potential to reveal cerebellar changes that precede functional outcomes. We applied this approach in subacute CS patients to characterize longitudinal structural trajectories that may provide mechanistic insight into post-stroke cognitive dysfunction. Hypothesis: Diffusion imaging metrics will reveal longitudinal microstructural alterations in CS. Methods: 14 subacute ischemic stroke participants (PPT) were imaged at 1 month and 6 months post-stroke onset on a 3T Siemens Prisma MRI scanner using an advanced multi-shell diffusion-weighted protocol. Diffusion tensor imaging (DTI) metrics included fractional anisotropy (FA), mean (MD), axial (AD), and radial diffusivity (RD), and non-Gaussian Diffusion (NGD) metric, complexity (C=1-α). PPT mean age was 65.4(11.5) years; 12 males; race (1 Latino/Hispanic, 2 Asian, 2 Black, 9 White). NIHSS median = 1 (IQR 0-3). Montreal Cognitive Assessment (MoCA) medians were 27 IQR 26-29 at 1M and 28 (IQR 27-29) at 6M. 3 PPTs had CS, and 11 PPTs had noncerebellar stroke (NCS). Analyses compared CS versus NCS participants for diffusion metrics in lesioned and contralesional cerebellar white matter for longitudinal change and examined associations with MoCA. Results: MoCA scores did not differ significantly between CS and NCS at 1M or 6M. The change from 1M to 6M trended toward significance (p=0.052), with CS showing improvement and NCS showing decline. Fig. 1 shows significant clusters of increased AD, RD, and MD in the infarcted hemisphere of CS compared to NCS (p<0.05). NGD analysis revealed a contralesional cluster (p<0.05) showing increased complexity in CS (Fig. 2). Conclusions: Longitudinal diffusion analysis revealed greater increases in diffusivity within affected cerebellar white matter in CS compared to NCS. Conventional DTI showed no contralesional changes, whereas NGD metrics identified increased complexity, suggesting evolving tissue heterogeneity and possible compensatory response. MoCA changes trended toward improvement in CS and decline in NCS, consistent with potential cognitive resilience paralleling contralesional remodeling. These findings underscore the sensitivity of advanced diffusion imaging to microstructural dynamics that may precede cognitive impairment.