Complex brain tissue compartment modelling using novel multi-shell diffusion techniques shows superior sensitivity in detecting microstructural changes and dynamic recovery post-stroke. Although limited evidence suggests that NODDI-derived parameters may predict motor recovery post-stroke, no studies to date have investigated NODDI parameters in relation to stroke cognitive outcomes. 

Multi-shell diffusion-weighted imaging and cognitive data from 63 patients at 3 months post-stroke were collected as part of the IC3 study (Imperial Comprehensive Cognitive assessment in Cerebrovascular disease^1,2). Twenty patients had repeat assessment at 1-year. NODDI-derived multi-compartment modelling was used to assess Neurite Density Index (NDI), Orientation Dispersion Index (ODI), and Isotropic Volume Fraction (ISO). These were correlated to detailed cognitive measures, after dimensionality reduction of IC3 scores using Principal Component Analysis (PCA).

Lesioned grey matter (GM) showed reduced ODI and NDI compared to healthy tissue(P<0.03). Similarly, NDI was reduced within lesioned white matter (WM) and WM hyperintensities(P<0.02). ODI decreased longitudinally in lesioned GM(P=0.001) and WM(P=0.028). Total WM NDI negatively correlated with modified Fazekas scores(R²=-0.2806,P<0.0001). 

Higher NDI within the WM and ODI within the anterior and posterior cingulate (ACC/PCC) positively correlated to the first PCA factor explaining 34% variance, largely mapping to language, attention, processing speed and verbal working memory. This aligns with cingulate’s complex role in cognitive processes regulation and ACC’s involvement in the multiple demand network. 

Microstructural changes in stroke-damaged tissue can be quantified using NODDI, showing decreased neurite density and simplified network complexity. More severe forms of small-vessel disease may lead to compromised WM microstructural integrity, correlating with poorer cognitive outcomes. 

¹Gruia,D.-C.et al.(2023)‘IC3 protocol:A longitudinal observational study of cognition after stroke using novel Digital Health Technology’,BMJOpen,13(11).

²Gruia,D.-C.et al.(2024)‘Online monitoring technology for deep phenotyping of cognitive impairment after stroke’,medRxiv,2024.09.06.24313173[Preprint].