Development of a stroke risk score with MRI asymptomatic brain lesions attributes to evaluate prognostic vascular events

Abstract

BackgroundThe use of a combination of stroke predictors, such as clinical factors and asymptomatic lesions on brain magnetic resonance imaging (MRI), may improve the accuracy of stroke risk prediction. Therefore, we attempted to develop a stroke risk score for healthy individuals. MethodsWe investigated the presence of cerebral stroke in 2365 healthy individuals who underwent brain dock screening at the Health Science Center in Shimane. We examined the factors that contributed to stroke and attempted to determine the risk of stroke by comparing background factors and MRI findings. ResultsThe following items were found to be significant risk factors for stroke: age (≥60 years), hypertension, subclinical cerebral infarction, deep white matter lesion, and microbleeds. Each item was scored with 1 point, and the hazard ratios for the risk of developing stroke based on the group with 0 points were 17.2 (95% confidence interval [CI] 2.31–128) for 3 points, 18.1 (95% CI 2.03–162) for 4 points, and 102 (95% CI 12.6–836) for 5 points. ConclusionsA precise stroke prediction score biomarker can be obtained by combining MRI findings and clinical factors.