A machine learning based death risk analysis and prediction of ST-segment elevation myocardial infarction (STEMI)...

Area of Science:

• Cardiology
• Medical Informatics
• Artificial Intelligence in Medicine

Background:

• ST-segment elevation myocardial infarction (STEMI) is a critical condition requiring rapid diagnosis and intervention.
• Traditional risk assessment for STEMI patients is often subjective and time-consuming.
• AI applications are increasingly vital for early diagnosis and treatment in modern medicine.

Purpose of the Study:

• To develop and evaluate machine learning models for predicting in-hospital mortality risk in STEMI patients.
• To provide clinical decision support for managing STEMI patients.
• To enhance the accuracy and speed of risk assessment compared to traditional methods.

Main Methods:

• Utilized selected machine learning algorithms including Support Vector Machines (SVM), Random Forest (RF), k-Nearest Neighbors (k-NN), Logistic Model Trees (LMT), and Multilayer Perceptrons (MLP).
• Applied these algorithms for death risk analysis and in-hospital mortality risk prediction in STEMI patients.
• Evaluated model performance using metrics such as accuracy, recall, precision, F-score, and Area Under the Curve (AUC).

Main Results:

• The proposed machine learning methods achieved superior performance, exceeding 99% in all evaluated metrics (accuracy, recall, precision, sensitivity, F-score, AUC).
• The models demonstrated high predictive power even with a reduced number of predictors.
• The AI-based approach outperformed existing studies in the literature for STEMI risk prediction.

Conclusions:

• Machine learning algorithms offer a powerful tool for accurate and rapid risk prediction in STEMI patients.
• The developed models can serve as effective clinical decision support systems, improving patient management and healthcare outcomes.
• AI-driven risk assessment can significantly enhance the efficiency and reliability of STEMI patient care.