Construction of a predictive model for patients with ischemic stroke based on clinical features of carotid plaques
10.19845/j.cnki.zfysjjbzz.2024.0167
- VernacularTitle:基于颈动脉斑块的临床特征构建缺血性卒中患者预测模型的研究
- Author:
Xinyi CAI
1
;
Pinjing HUI
;
Yanhong YAN
Author Information
1. 苏州大学附属第一医院神经外科-颈脑血管超声科,江苏 苏州 215006;苏州大学附属第一医院卒中中心,江苏 苏州 215006
- Keywords:
Ischemic stroke;
Carotid;
Atherosclerotic plaque;
Inflammation;
Predictive model
- From:
Journal of Apoplexy and Nervous Diseases
2024;41(10):875-881
- CountryChina
- Language:Chinese
-
Abstract:
Objective To construct a predictive model for patients with ischemic stroke(IS)based on the clinical features of carotid plaques.Methods A total of 189 patients with suspected acute IS who were admitted to the Depart-ment of Neurology at the First Affiliated Hospital of Soochow University for the first time were included in the study.Ac-cording to the results of diffusion-weighted imaging(DWI)of cranial magnetic resonance imaging,these patients were di-vided into IS group and non-ischemic stroke(NIS)group.According to the features of carotid atherosclerotic plaque as-sessed by ultrasound,the lesions were divided into vulnerable plaques and stable plaques.Logistic regression was used to analyze the nature of carotid plaques and clinical features in patients with IS,and the area under the ROC curve was calcu-lated.Results There were significant differences in the number of patients with vulnerable plaques,degree of carotid ste-nosis,hypertension,diabetes,and neutrophil-lymphocyte ratio(NLR)between the IS group and the NIS group(P<0.05).Multivariate logistic regression analysis revealed that the presence of vulnerable carotid plaques,poor control of hypertension and diabetes,and elevated NLR were the main influencing factors for IS events(P<0.05).The area under the ROC curve of vulnerable carotid plaques combined with clinical features(poor control of hypertension and diabetes and elevated NLR)to predict the occurrence of IS events in patients was 0.776(95%confidence interval:0.709?0.842),with a sensitivity of 0.748 and a specificity of 0.758.Conclusion The predictive model based on vulnerable carotid plaques combined with clinical features shows promise.
