Early Diagnoses of Acute Coroner Syndrome Based on Machine Learning Model

Abstract

Cardiovascular diseases are a leading global cause of death, particularly in low to middle-income countries. Early and accurate diagnosis of Acute Coronary Syndrome (ACS) is vital, but limited access to healthcare hinders effective management. This study utilized machine learning to develop mathematical models for ACS risk detection. Data from 249 individuals with ACS or suspected heart disease were used to construct twelve models with different parameters and classifiers. Performance indicators, including accuracy, Matthews correlation coefficient, and precision, were employed for evaluation. The Random Forest classifier demonstrated superior performance, achieving 90.45% accuracy for internal validation and 86% for external validation. Critical criteria for ACS diagnosis were CK-MB, age, coronary artery disease, and Troponin T value. The models developed in this study significantly prevent potential deaths via rapid intervention and reduce healthcare expenditures by minimizing unnecessary human resources and repeat tests.