Objective: White matter hyperintensity (WMH) is an important imaging feature of cerebral small vessel disease (CSVD). In this study, a data-driven method was used to conduct quantitative analysis of WMH, which was compared with the traditional semi-quantitative assessment, to explore the correlation between WMH and early cognitive impairment in post-stroke patients with CSVD.

Methods: From July 2015 to February 2018, 117 post-stroke patients with CSVD who met the selection criteria were recruited consecutively from the Stroke Clinic of Renji Hospital, Shanghai Jiao Tong University School of Medicine, comprising 72 individuals with mild cognitive impairment (MCI) and 45 with no cognitive impairment (NCI). There were no significant differences in baseline demographic sociology and vascular risk factors between the two groups (P > 0.05). Detailed neuropsychological assessments and multimodal magnetic resonance imaging (MRI) were performed in all patients. The WMH was semiquantitatively scored by Fazekas scale and the volume of WMH was quantitatively calculated. The cognitive function was compared between the two groups, and the correlation between WMH semi-quantitative Fazekas score and quantitative score of WMH volume and cognitive function was analyzed.

Results: There were significant differences in cognitive domains including attentionexecution, visual space, language and memory functions between MCI and NCI patients (P < 0.05), but there was no significant difference in WMH semi-quantitative Fazekas score (P = 0.090), while the quantitative score of WMH volume was significant different between MCI and NCI patients (P = 0.004). After adjusting for age, gender and years of education, there was no correlation between WMH semi-quantitative Fazekas score and cognitive functions (P > 0.05), but quantitative score of WMH volume was significantly correlated with cognitive function (P < 0.05). Deep WMH volume was significantly correlated with attentionexecution and visual space functions (P < 0.05), while the periventricular WMH volume was significantly correlated with cognitive functions (P < 0.05).

Conclusion: WMH is closely related to cognitive decline in post-stroke patients with CSVD. Quantitative score of WMH volume is more sensitive than WMH semi-quantitative Fazekas score in assessing early cognitive function changes in patients with CSVD, and the relationship of WMH location with cognitive fuction exhibits heterogeneity.