Abstract: Objective To explore influencing factors of anterior circulation ischemic stroke in patients with carotid plaque by the clinical and multimodal ultrasound (MMU), and to construct a risk stratification prediction model based on superb microvascular imaging (SMI).
Methods A total of 683 patients with carotid plaque were retrospectively analyzed. They were assigned to anterior circulation stroke group(n=301) and non-anterior circulation stroke group(n=382) based on clinical presentation and computed tomography (CT)/magnetic resonance imaging (MRI). MMU characteristics of carotid plaques, and clinical and laboratory examination data were collected. Multivariate binary Logistic regression analysis was used to screen the influencing factors of anterior circulation stroke. The nomogram prediction model was constructed, to carry out model verification and risk stratification. 
Results There were statistically significant differences  in age, body mass index（BMI）, drinking history, smoking history, previous history of cerebral infarction, hypertension, diabetes, low-density lipoprotein （LDL）, high-density lipoprotein （HDL）, homocysteine （HCY）, carotid stenosis, plaque surface morphology and intraplaque neovascularization between anterior circulation stroke group and non-anterior circulation stroke group (P＜0.05).Multivariate binary Logistic regression analysis found that age, BMI, smoking history, drinking history, previous history of cerebral infarction, diabetes, LDL, HCY and MMU (carotid stenosis, plaque surface morphology, intraplaque neovascularization) were independent risk factors for anterior circulation stroke in patients with carotid plaque(P＜0.05). Based on the above 11 indicators, the individualized nomogram prediction model of anterior circulation stroke was constructed. The total score of the obtained nomogram could more effectively predict the risk of anterior circulation stroke in patients with carotid plaque (AUC: 0.781,95%CI: 0.747-0.816， Hosmer-lemeshow P=0.637). The optimal cutoff value of the model was determined to be 0.465(the nomogram score:138), which divided patients with carotid plaque into low-risk and high-risk subgroups. The clinical performance of the model was verified by Bootstrap. 
Conclusion The prediction model based on MMU can readily, quickly and accurately predict the occurrence and risk stratification of anterior circulation stroke in patients with carotid plaque.

Key words: ischemic stroke, carotid stenosis, multimodal ultrasound