A stacking-based model for predicting 30-day all-cause hospital readmissions of patients with acute myocardial infarction
Abstract Background Acute myocardial infarction (AMI) is a serious cardiovascular disease, followed by a high readmission rate within 30-days of discharge. Accurate prediction of AMI readmission is a crucial way to identify the high-risk group and optimize the distribution of medical resources. Meth...
Main Authors: | , , , |
---|---|
Format: | Article |
Language: | English |
Published: |
BMC
2020-12-01
|
Series: | BMC Medical Informatics and Decision Making |
Subjects: | |
Online Access: | https://doi.org/10.1186/s12911-020-01358-w |
id |
doaj-1a4ae28e2bb042c0bd8140de33986264 |
---|---|
record_format |
Article |
spelling |
doaj-1a4ae28e2bb042c0bd8140de339862642020-12-20T12:35:42ZengBMCBMC Medical Informatics and Decision Making1472-69472020-12-0120111310.1186/s12911-020-01358-wA stacking-based model for predicting 30-day all-cause hospital readmissions of patients with acute myocardial infarctionZhen Zhang0Hang Qiu1Weihao Li2Yucheng Chen3School of Computer Science and Engineering, University of Electronic Science and Technology of ChinaSchool of Computer Science and Engineering, University of Electronic Science and Technology of ChinaCardiology Division, West China Hospital, Sichuan UniversityCardiology Division, West China Hospital, Sichuan UniversityAbstract Background Acute myocardial infarction (AMI) is a serious cardiovascular disease, followed by a high readmission rate within 30-days of discharge. Accurate prediction of AMI readmission is a crucial way to identify the high-risk group and optimize the distribution of medical resources. Methods In this study, we propose a stacking-based model to predict the risk of 30-day unplanned all-cause hospital readmissions for AMI patients based on clinical data. Firstly, we conducted an under-sampling method of neighborhood cleaning rule (NCR) to alleviate the class imbalance and then utilized a feature selection method of SelectFromModel (SFM) to select effective features. Secondly, we adopted a self-adaptive approach to select base classifiers from eight candidate models according to their performances in datasets. Finally, we constructed a three-layer stacking model in which layer 1 and layer 2 were base-layer and level 3 was meta-layer. The predictions of the base-layer were used to train the meta-layer in order to make the final forecast. Results The results show that the proposed model exhibits the highest AUC (0.720), which is higher than that of decision tree (0.681), support vector machine (0.707), random forest (0.701), extra trees (0.709), adaBoost (0.702), bootstrap aggregating (0.704), gradient boosting decision tree (0.710) and extreme gradient enhancement (0.713). Conclusion It is evident that our model could effectively predict the risk of 30-day all cause hospital readmissions for AMI patients and provide decision support for the administration.https://doi.org/10.1186/s12911-020-01358-wAcute myocardial infarctionHospital readmissionClinical dataMachine learningSelf-adaptiveStacking-based model learning |
collection |
DOAJ |
language |
English |
format |
Article |
sources |
DOAJ |
author |
Zhen Zhang Hang Qiu Weihao Li Yucheng Chen |
spellingShingle |
Zhen Zhang Hang Qiu Weihao Li Yucheng Chen A stacking-based model for predicting 30-day all-cause hospital readmissions of patients with acute myocardial infarction BMC Medical Informatics and Decision Making Acute myocardial infarction Hospital readmission Clinical data Machine learning Self-adaptive Stacking-based model learning |
author_facet |
Zhen Zhang Hang Qiu Weihao Li Yucheng Chen |
author_sort |
Zhen Zhang |
title |
A stacking-based model for predicting 30-day all-cause hospital readmissions of patients with acute myocardial infarction |
title_short |
A stacking-based model for predicting 30-day all-cause hospital readmissions of patients with acute myocardial infarction |
title_full |
A stacking-based model for predicting 30-day all-cause hospital readmissions of patients with acute myocardial infarction |
title_fullStr |
A stacking-based model for predicting 30-day all-cause hospital readmissions of patients with acute myocardial infarction |
title_full_unstemmed |
A stacking-based model for predicting 30-day all-cause hospital readmissions of patients with acute myocardial infarction |
title_sort |
stacking-based model for predicting 30-day all-cause hospital readmissions of patients with acute myocardial infarction |
publisher |
BMC |
series |
BMC Medical Informatics and Decision Making |
issn |
1472-6947 |
publishDate |
2020-12-01 |
description |
Abstract Background Acute myocardial infarction (AMI) is a serious cardiovascular disease, followed by a high readmission rate within 30-days of discharge. Accurate prediction of AMI readmission is a crucial way to identify the high-risk group and optimize the distribution of medical resources. Methods In this study, we propose a stacking-based model to predict the risk of 30-day unplanned all-cause hospital readmissions for AMI patients based on clinical data. Firstly, we conducted an under-sampling method of neighborhood cleaning rule (NCR) to alleviate the class imbalance and then utilized a feature selection method of SelectFromModel (SFM) to select effective features. Secondly, we adopted a self-adaptive approach to select base classifiers from eight candidate models according to their performances in datasets. Finally, we constructed a three-layer stacking model in which layer 1 and layer 2 were base-layer and level 3 was meta-layer. The predictions of the base-layer were used to train the meta-layer in order to make the final forecast. Results The results show that the proposed model exhibits the highest AUC (0.720), which is higher than that of decision tree (0.681), support vector machine (0.707), random forest (0.701), extra trees (0.709), adaBoost (0.702), bootstrap aggregating (0.704), gradient boosting decision tree (0.710) and extreme gradient enhancement (0.713). Conclusion It is evident that our model could effectively predict the risk of 30-day all cause hospital readmissions for AMI patients and provide decision support for the administration. |
topic |
Acute myocardial infarction Hospital readmission Clinical data Machine learning Self-adaptive Stacking-based model learning |
url |
https://doi.org/10.1186/s12911-020-01358-w |
work_keys_str_mv |
AT zhenzhang astackingbasedmodelforpredicting30dayallcausehospitalreadmissionsofpatientswithacutemyocardialinfarction AT hangqiu astackingbasedmodelforpredicting30dayallcausehospitalreadmissionsofpatientswithacutemyocardialinfarction AT weihaoli astackingbasedmodelforpredicting30dayallcausehospitalreadmissionsofpatientswithacutemyocardialinfarction AT yuchengchen astackingbasedmodelforpredicting30dayallcausehospitalreadmissionsofpatientswithacutemyocardialinfarction AT zhenzhang stackingbasedmodelforpredicting30dayallcausehospitalreadmissionsofpatientswithacutemyocardialinfarction AT hangqiu stackingbasedmodelforpredicting30dayallcausehospitalreadmissionsofpatientswithacutemyocardialinfarction AT weihaoli stackingbasedmodelforpredicting30dayallcausehospitalreadmissionsofpatientswithacutemyocardialinfarction AT yuchengchen stackingbasedmodelforpredicting30dayallcausehospitalreadmissionsofpatientswithacutemyocardialinfarction |
_version_ |
1724376341094072320 |