Ischemic Heart Disease Prediction Frameworks Using Machine Learning: A Review

Ischemic heart disease (IHD) remains a leading cause of death worldwide, necessitating robust predictive models for early intervention. Traditional risk assessments lack the precision required for timely detection and prevention, highlighting the critical need for advanced machine learning (ML) models. This review comprehensively examines recent advancements in ML-based IHD prediction models. It focuses on supervised, unsupervised, and reinforcement learning techniques, detailing their objectives, performance metrics, limitations, and datasets. The analysis reveals that Random Forest and Support Vector Machines are frequently adopted due to their high accuracy, often exceeding 90%. Despite these advancements, challenges such as small, imbalanced datasets and data privacy concerns persist. Trends indicate a growing emphasis on utilizing comprehensive datasets and personalized approaches to enhance model reliability and clinical applicability. Additionally, there is an increased focus on feature selection and data quality to improve prediction accuracy. This review highlights the need for collaboration between healthcare institutions and data scientists to improve data acquisition and model interpretability. By addressing these challenges, future research can develop robust, explainable ML tools that significantly impact public health by enabling earlier and more accurate IHD detection. The insights from this review aim to guide researchers in creating effective ML models that can mitigate the global burden of IHD and enhance patient outcomes through early intervention and precise risk assessment.