Association Between Cognitive Impairment and Multimodal MRI Markers in Patients with Cerebral Microbleeds: A Retrospective Case-Control Study.

OBJECTIVE

To investigate the association between cognitive impairment and multimodal magnetic resonance imaging (MRI) markers in patients with cerebral microbleeds (CMBs), and to identify imaging predictors of CMB-related cognitive dysfunction.

METHODS

This retrospective case-control study included 71 patients with CMBs confirmed by MRI between August 2022 and February 2024 and 65 age-matched healthy controls. Cognitive function was assessed using the Montreal Cognitive Assessment (MoCA). All participants underwent multimodal MRI, including susceptibility-weighted imaging (SWI) for CMB detection and diffusion tensor imaging (DTI) for quantitative assessment of fractional anisotropy (FA) and apparent diffusion coefficient (ADC) in predefined brain regions. Spearman correlation analysis was performed to evaluate associations between CMB burden, DTI parameters, and cognitive scores. Multivariate linear regression analysis was used to identify independent risk factors for cognitive impairment.

RESULTS

Compared with controls, patients with CMBs had significantly lower MoCA scores and higher CMB burden (both P < 0.05). FA values in the frontal-temporal lobe, parietal lobe, and basal ganglia were significantly reduced, while ADC values were increased (all P < 0.05). CMB number was negatively correlated with MoCA scores (r = -0.643, P < 0.001). Decreased FA and increased ADC in the frontal-temporal lobe and basal ganglia were significantly associated with cognitive decline. Regression analysis showed that CMB burden ≥10, reduced FA, elevated ADC, and basal ganglia involvement were independent risk factors for cognitive impairment.

CONCLUSION

Cognitive impairment in CMB patients is closely associated with lesion burden and microstructural white matter alterations detected by multimodal MRI. These imaging markers may facilitate early risk stratification and targeted monitoring, although longitudinal validation is warranted.