A Stacking Ensemble Prediction Model for the Occurrences of Major Adverse Cardiovascular Events in Patients With Acute Coronary Syndrome on Imbalanced Data

Abstract

The major adverse cardiovascular events (MACE) often occur with high morbidity and mortality globally. It is very important to predict the MACE occurrences accurately in patients with acute coronary syndrome (ACS). Therefore, this paper proposes a stacking ensemble model for the prediction of MACE occurrences in patients with ACS at early stage. Our research contents are given as follows. First, we use Korea Acute Myocardial Infarction Registry National Institutes of Health (KAMIR-NIH) dataset and experimental data are extracted from the raw data and preprocessed. Second, we apply three data sampling approaches, such as borderline synthetic minority oversampling technique (Borderline-SMOTE1), cluster centroids undersampling, and synthetic minority oversampling techniques (SMOTE) plus Tomek Links (SMOTETomek) hybrid technique, to solve the class imbalance problem. Third, to develop a stacking ensemble prediction model for the occurrences of MACE, we apply seven widely used machine learning algorithms, such as logistic regression (LR), support vector machine (SVM), K-Nearest Neighbors (KNN), decision tree (DT), random forest (RF), extreme gradient boosting (XGBoost), and adaptive boosting (AdaBoost), as base learners. Fourth, the performance of proposed stacking ensemble model is compared with the seven base learners using the three data sampling techniques. In the result, the proposed stacking ensemble model with the SMOTETomek showed the best performance with 0.9862 accuracy, 0.9976 precision, 0.975 recall, 0.9862 f1_score, 0.9863 g-mean, and 0.9863 AUC and provided a better solution for imbalanced dataset. Consequently, our finding was that the proposed stacking ensemble model with the SMOTETomek outperformed the base learners and improved the accuracy of diagnosis and prediction of the MACE occurrences in patients with ACS at early stage.INDEX TERMS Major adverse cardiovascular events (MACE), acute coronary syndrome (ACS), stacking ensemble classifier, machine learning, data sampling, imbalanced data.

Analysis of the blood bacterial composition of patients with acute coronary syndrome and chronic coronary syndrome

“…It emphasizes that the stability of non-acute coronary heart disease is only relative, and there is a risk of progression to Acute coronary syndrome at any time, which leads to cardiovascular events ( Steg et al., 2007 ). Acute coronary syndrome is also a common subcategory of cardiovascular disease and has led to increased mortality globally ( Zheng et al., 2021 ). Acute coronary syndrome is a set of signs and symptoms due to acutely decreased blood flow in the coronary arteries, and the exact mechanism underlying its pathogenesis remains to be fully elucidated ( Dai et al., 2020 ).…”

A Machine Learning-Based Applied Prediction Model for Identification of Acute Coronary Syndrome (ACS) Outcomes and Mortality in Patients during the Hospital Stay

“…Data sampling is an efficient and valuable technique for the transformation of an imbalanced training dataset into balanced class distribution either by increasing the minority class data (oversampling), or decreasing the majority class data (undersampling), or using both strategies at the same time (hybrid sampling). From previous studies of health informatics research and data sampling techniques [ 39 , 40 , 41 , 42 ], it has been concluded that hybrid sampling techniques are considered the best data sampling techniques. According to our target variables, our final prediction results include the ACS outcomes along with the discharge reason and death type (final diagnosis + discharge result + death type) for the patients admitted to hospital due to heart problems.…”

Pre-hospital prediction of adverse outcomes in patients with suspected COVID-19: Development, application and comparison of machine learning and deep learning methods

“…The meta-learner is then trained to optimally combine the predictions of base-learners , to form a final set of predictions. An example of its use in medical research is to predict occurrences of major adverse cardiovascular events in patients with acute coronary syndromes [33] . A simplified diagram of stacking is shown in Fig.…”