Clinical Correlation of Magnetic Resonance Imaging Features of Cerebral Small Vessel Disease

Abstract

Objective: We have discussed the clinical features and neuroimaging findings and investigated the correlations between the clinical characteristics and different neuroradiologic phenotypes of cerebral small vessel disease (CSVD).

Methods: A total of 1106 patients with a diagnosis of CSVD were enrolled. The demographic data, medical history, laboratory test results, and neuroimaging findings were retrieved for analysis. The differences in clinical parameters between patient groups were examined. The relationships between the clinical parameters and neuroradiologic phenotypes (i.e., white matter lesions [WMLs] and enlarged perivascular spaces) were assessed. The magnetic resonance imaging features were clustered using the fast clustering algorithm.

Results: Approximately one third of our patients presented with a lacune, which was associated with atrial fibrillation (P = 0.029), lacunar syndrome (P < 0.001), periventricular WMLs (P = 0.001), cerebral WMLs (P = 0.021), basal ganglia perivascular space grade (P < 0.001) and severity (P = 0.001), and semiovale perivascular space grade (P = 0.010) and severity (P = 0.002). Hypertension was associated with periventricular WMLs (P = 0.048), centrum semiovale WMLs (P = 0.026), and basal ganglia perivascular space grade (P < 0.001) and severity (P < 0.001). A novel clustering algorithm was derived to stratify our cohort into 3 different groups according to the differing severity of the cerebral WMLs and perivascular space enlargement.

Conclusion: The present study has provided a comprehensive analysis of the clinical correlation of characteristics and neuroradiologic phenotypes in patients with CSVD. The insights from these findings could be used to refine the management strategy for patients with CSVD.

Keywords: Basal ganglia perivascular space; Centrum semiovale white matter lesion; Cerebral small vessel disease; Periventricular white matter lesion.