Article Detail

Ömer Utku Kılınç*

Keywords

Myocardial Infarction Machine Learning Clinical Decision Support Systems XGBoost RFECV Feature Selection.

Doi : 10.71350/hgtsci.2026165

Abstract

Myocardial infarction (MI), commonly known as a heart attack, is a life-threatening health condition that causes tissue damage due to reduced blood flow to the heart muscle. This study aimed to predict lethal outcomes using clinical data at the time of hospital admission, based on a dataset of 1,700 patient records collected by the Krasnoyarsk Interdistrict Clinical Hospital between 1992 and 1995. Contrary to studies in the literature, this study followed a rigorous methodology that minimized the risk of “data leakage”; complication columns were removed from the pre-training dataset, and feature selection (RFECV) and modeling processes were isolated from each other. Various algorithms, such as XGBoost, CatBoost, Random Forest, SVM, and MLP, were tested with different parameter combinations (e.g., tree depth, number of estimators). The results show that the XGBoost algorithm, with parameters max_depth=7 and n_estimators=150, achieved the highest success rate, reaching an F1 Score of 83.94% and an Accuracy rate of 87.65%. This study demonstrates that class weighting and hybrid feature selection methods yield more reliable results compared to synthetic data generation (SMOTE) in imbalanced medical datasets.

References

1. Elashmawi, W. H., Tamer, A., & Nasr, A. (2024). Multi-target classification for predicting myocardial infarction complications: A comparative analysis of various machine learning techniques. 2024 International Mobile, Intelligent, and Ubiquitous Computing Conference (MIUCC). 10.1109/MIUCC62295.2024.10783524
3. Faizal, A. S. M., Hon, W. Y., Thevarajah, T. M., Low, T. Y., & Ibrahim, F. (2023). A biomarker discovery of acute myocardial infarction using feature selection and machine learning. Medical & Biological Engineering & Computing. https://doi.org/10.1007/s11517-023-02832-w
5. Ghafari, R., Azar, A. S., Ghafari, A., & Kabir, A. (2023). Prediction of the fatal acute complications of myocardial infarction via machine learning algorithms. The Journal of Tehran University Heart Center. https://doi.org/10.18502/jthc.v18i4.14515
7. Hamidi, H., & Daraei, A. (2018). A novel two-step feature selection based cost sensitive myocardial infarction prediction model. International Journal of Computational Intelligence Systems. https://doi.org/10.2991/ijcis.11.1.66
9. Khamis, G. S. M., Mohammed, Z. M. S., Alanazi, S. M., Ali, A. M., & Al-shammari, A. S. (2024). Prediction of myocardial infarction complications using Gradient Boosting. Engineering, Technology & Applied Science Research. https://doi.org/10.48084/etasr.8415
11. Newaz, A., Mohosheu, M. S., & Al Noman, M. A. (2023). Predicting complications of myocardial infarction within several hours of hospitalization using data mining techniques. Informatics in Medicine Unlocked. https://doi.org/10.1016/j.imu.2023.101361
13. Satty, A., Salih, M. M. Y., Hassaballa, A. A., & Ahmed, M. E. (2024). Comparative analysis of machine learning algorithms for investigating myocardial infarction complications. Engineering, Technology & Applied Science Research. https://doi.org/10.48084/etasr.6482
15. Sulthana, A. R., & Jaithunbi, A. K. (2022). Varying combination of feature extraction and modified support vector machines based prediction of myocardial infarction. Evolving Systems. https://doi.org/10.1007/s12530-021-09411-0
17. Wang, Y., Shen, J., Li, S., & Zhang, X. (2024). Evaluating the effectiveness of feature selection and explainable AI in predicting acute myocardial infarction using machine learning models. 2024 5th International Conference on Machine Learning and Computer Application (ICMLCA). 10.1109/ICMLCA63499.2024.10753731