Construction and Validation of A Prediction Model Based on Multiple Blood Biomarkers for the Differential Diagnosis of Acute Aortic Dissection and Acute Myocardial Infarction
10.3969/j.issn.1671-7414.2025.03.033
- VernacularTitle:构建多种血液标志物预测模型用于急性主动脉夹层与急性心肌梗死的鉴别诊断及验证
- Author:
Xinyu HUANG
1
;
Shiqiang XIANG
Author Information
1. 武汉科技大学医学院公共卫生学院,武汉 430065;华中科技大学同济医学院附属同济医院检验科,武汉 430030
- Publication Type:Journal Article
- Keywords:
acute aortic dissection;
acute myocardial infarction;
blood biomarkers;
prediction model;
least absolute shrinkage and selection operater regression
- From:
Journal of Modern Laboratory Medicine
2025;40(3):177-182
- CountryChina
- Language:Chinese
-
Abstract:
Objective To differentiate acute aortic dissection(AAD)from acute myocardial infarction(AMI)using commonly used blood biomarkers and develop and validate a diagnostic model.Methods A total of 350 chest pain patients presenting to Tongji Hospital,Tongji Medical College,Huazhong University of Science and Technology,from January to December,were selected,including 200 AAD patients and 150 AMI patients.The patients were randomly divided into training and validation set in a 4∶3 ratio.Nineteen blood biomarker values were collected from all patients,and the least a bsolute shrinkage and selection operator(LASSO)regression model was applied to the training set to identify the optimal predictors for constructing a differential diagnostic model between AAD and AMI.The diagnostic efficacy of the model was assessed using receiver operating characteristic(ROC)curves.Results Compared with the AMI group,patients in the AAD group exhibited significantly higher levels of monocyte count(MONO),neutrophil count(NEU),high-sensitivity C-reactive protein(hs-CRP),high-density lipoprotein(HDL),prothrombin time(PT)and D-dimer(D-D)(U=17 160～26 593,all P<0.05),whereas levels of lymphocyte count(LYM),triacylglycerol(TG),activated partial thromboplastin time(APTT),thrombin time(TT),fibrinogen(FIB),high-sensitivity cardiac troponin I(hs-cTnI),N-terminal pro-brain natriuretic peptide(NT-proBNP),aspartate aminotransferase(AST)and creatine kinase(CK)were significantly lower(U=8 943～13 063,all P<0.05).LASSO regression analysis identified D-D,hs-cTnI,NEU,LYM,hs-CRP and APTT as optimal predictors in the training set(U=2 433～8 657,all P<0.05).The constructed diagnostic model was as follows:Y=0.101(D-D)+0.065(NEU)+0.003(hs-CRP)-0.200(LYM)-0.004(APTT)-0.000 046 6(hs-cTnI)-0.006,with an optimal cutoffvalue of 0.093.The AUC(95%CI)was 0.937(0.90～0.97,P<0.01)in the training set,0.90(0.85～0.95,P<0.05)in the validation set,and 0.925(0.90～0.95,P<0.01)in the overall dataset,with a diagnostic sensitivity of 87.9%and specificity of 92.3%.Conclusion The diagnostic model based on D-D,hs-cTnI,NEU,LYM,hs-CRP,and APTT provides a valuable reference for the differential diagnosis of AAD and AMI.
