Loading...

Prediction of Myocardial Infarction Complications using Machine Learning Methods

Zojaji, Sahar | 2025

0 Viewed
  1. Type of Document: M.Sc. Thesis
  2. Language: Farsi
  3. Document No: 58630 (01)
  4. University: Sharif University of Technology
  5. Department: Industrial Engineering
  6. Advisor(s): Rafiee, Majid; Hemmati, Soheil
  7. Abstract:
  8. Accurate prediction of myocardial infarction complications is one of the major challenges in the management of cardiac patients, as the occurrence of such complications can lead to severe consequences for quality of life, length of hospitalization, treatment costs, and mortality. This study aims to develop a data-driven, machine learning-based framework to predict 12 critical post-infarction outcomes using modern algorithms. In the first step, patients were stratified into more homogeneous groups based on treatment effects through a causal clustering algorithm. Subsequently, machine learning models, including logistic regression, random forest, and gradient boosting, were trained and evaluated both within each cluster and on the overall dataset. To assess model performance, accuracy, the area under the curve (AUC), and the Scott-Knott statistical ranking test were employed. The results indicated that in many outcomes, models incorporating causal clustering achieved superior performance compared to global models, with improvements that were not only numerical but also statistically significant. These findings present a novel approach for analyzing patient heterogeneity and modeling cardiac complications, offering the potential to serve as a foundation for developing data-driven clinical decision support systems and health policy planning
  9. Keywords:
  10. Myocardial Infarction ; Machine Learning ; Complication Prediction ; Causal Clustering ; Heart Attack ; Cardiovascular Patients

 Digital Object List

 Bookmark

...see more