Explainable prediction model of in-hospital major adverse cardiovascular events in acute ST-segment elevation myocardial infarction based on machine learning
WANG Zheng
ZHANG Ting
YANG Lin-fei
XU Ming-sheng
ZHANG Jian
ZHANG Jing
Abstract：Objective To establish a prediction model for in-hospital major adverse cardiovascular events(MACE)in acute ST-segment elevation myocardial infarction(STEMI).Methods A total of 514 STEMI patients admitted to the Second People's Hospital of Hefei from June 2019 to June 2023 were retrospectively analyzed.The patients were divided into two groups according to whether MACE occurred or not.Three machine learning algorithms,namely Logistic regression(LR),support vector machine(SVM)and random forest(RF)were used to construct the prediction model for in-hospital MACE in STEMI patients.The prediction performance of the models was compared by sensitivity,specificity,accuracy,Fl score,area under the receiver operating characteristic curve(AUC)and calibration curve.The DALEX algorithm was used to visualize the importance ranking of risk factors in the best model.Results 23.15％of STEMI patients developed MACE.The LR prediction model had the best performance with an AUC of 83.9％.DALEX showed that Killip grade,homocysteine,urea,creatinine and admission systolic blood pressure were the top five risk factors for MACE in STEMI patients.Conclusion The LR machine learning algorithm has good efficacy in predicting in-hospital MACE in STEMI patients.Killip grade,homocysteine,urea,creatinine and admission systolic blood pressure are the important risk factors.
Keywords：Acute ST-segment elevation myocardial infarctionMajor adverse cardiovascular eventsMachine learningPredictive model
Publication Date：2025-02-09
Online Publishing Date：2025-08-15(First online date of this platform, not the publication date of the document)
Pages：6( 115-120 )