A Sustainable Approach to Predict Coronary Artery Disease with Machine Learning and Explainable AI

Coronary Artery Disease (CAD) is one of the major health concerns around the world. In this article, an effective and explainable CAD diagnosis system based on machine learning and Explainable AI methods is introduced. Strong pre-processing, cross-validation, and performance evaluation were employed on the Z-Alizadeh Sani dataset using nine various ML algorithms. Random Forest and Multi-Layer Perceptron models had the best results with 95.61% and 96.44% accuracies, respectively. This research study includes the use of CodeCarbon to reduce energy usage while training models to make it sustainable. SHAP analysis was also used to make it interpretable, hence making it clinically applicable. The uniqueness of this study is the combination of machine learning with sustainability-oriented AI to improve accuracy in diagnosis while minimizing environmental footprint, hence helping in the creation of effective and ethical AI-based health solutions.