Construction of restenosis risk warning model based on SMOTE algorithm after cerebrovascular intervention
10.19845/j.cnki.zfysjjbzz.2022.0185
- VernacularTitle:基于SMOTE算法的脑血管介入术后再狭窄风险预警模型的构建
- Author:
Yu LEI
1
;
Jiacai ZUO
1
Author Information
1. Mianyang Central Hospital,School of Medicine,University of Electronic Science and Technology of China,Mianyang 621000,China
- Publication Type:Journal Article
- Keywords:
Cerebrovascular disease;
Interventional therapy;
Risk factors;
Regression equation;
SMOTE algorithm
- From:
Journal of Apoplexy and Nervous Diseases
2022;39(8):741-745
- CountryChina
- Language:Chinese
-
Abstract:
Objective To explore the influencing factors of restenosis risk after cerebral vascular intervention,and establish a warning model based on SMOTE algorithm.Methods A total of 320 patients with cerebrovascular stenosis admitted to the Department of Neurology of Mianyang Central Hospital from May 2017 to May 2021 were selected.The medical records of all patients were retrospectively analyzed.According to whether restenosis occurred after interventional treatment,the patients were divided into stenosis group and non-stenosis group.Single factor and multi-factor Logistic regression analysis were used to screen the risk factors of restenosis after cerebrovascular intervention and establish the prediction model.At the same time,expand the positive group data based on SMOTE algorithm,build the early warning model of improved data set and compare and verify the prediction efficiency of the model.Results Smoking history,hyperlipidemia,diabetes mellitus,CRP≥5 mg/L and stent length≥16 mm were independent risk factors for restenosis after interventionotherapy in patients with cerebrovascular diseases (P<0.05).Based on the above risk factors,the AUC of early warning model P-1 and P-2 was established to be 0.872 (95%CI 0.821~0.923) and 0.847 (95%CI0.816~0.879),respectively.There was no significant difference in the efficacy of the two prediction models and their AUC was over 0.75,indicating that the prediction model had high efficacy.Conclusion Based on the history of smoking,hyperlipidemia,diabetes,CRP,bracket and the length of the original data of sampling algorithm to establish early warning model has higher predictive,medical personnel will enable effective intervention,thus reduce the patients with cerebrovascular disease risk for restenosis after interventional treatment,improve the prognosis of patients,improve the quality of survival.
- Full text:2024072323023908690Construction of restenosis risk warning model based on SMOTE algorithm after cerebrovascular intervention.pdf
