Artificial Intelligence for Early Heart Disease Prediction: A Review of Machine Learning Techniques

Abstract

Cardiovascular disease (CVD) is still the number one cause of death worldwide and many patients are exposed to severe cardiac events only after the disease has progressed to an advanced stage. Recent advances in artificial intelligence (AI) and machine learning (ML) have shown great potential in improving the early prediction of cardiovascular risk from electronic health records, physiological measurements and other clinical data. This paper provides an analytical review of recent studies on ML-based approaches for early heart disease and cardiogenic shock prediction. The study evaluates the performance of popular algorithms including Logistic Regression, Support Vector Machines, Random Forests, Gradient Boosting Machines, and neural networks. It also investigates the effect of data preparation techniques such as feature scaling, normalisation, and class balancing on prediction outcomes.Results show that ML models are superior to traditional risk score methods in terms of accuracy and can detect high risk patients much earlier than traditional clinical practice. However, data heterogeneity, missing data, model interpretability, and limited clinical validation continue to pose challenges for broad implementation, despite these promising results. The study concludes that there is an urgent need for explainable, clinically validated and standardised ML frameworks to translate predictive models into routine healthcare practice and improve early detection of cardiovascular disease. Keywords: artificial intelligence; machine learning; Electronic Health Records; cardiovascular disease; Early Disease Prediction