Explainable Machine Learning Framework for Cardiovascular Disease Diagnosis and Prognosis

Machine learning can significantly improve heart disease detection: The Random Forest model achieved 97.2% accuracy in identifying heart disease cases, demonstrating that AI-based approaches may help address diagnostic gaps, particularly in areas with limited access to healthcare infrastructure.

Explainable AI is essential for clinical adoption: The researchers specifically incorporated Explainable AI methods to make model predictions understandable to users, recognizing that doctors and patients need to understand why a diagnosis or risk prediction was made, not just what it is.

Synthetic data can help address real-world imbalances: The study used SMOTE to create 100,000 synthetic samples from an original dataset of 1,035 cases, demonstrating how AI systems can be trained more robustly when real-world medical data is limited or imbalanced—a common challenge in healthcare.

Different AI models serve different purposes: The research used both classification models (to detect disease presence) and regression models (to forecast risk levels), showing that responsible AI implementation requires selecting appropriate tools for specific clinical questions rather than applying one-size-fits-all solutions.

Performance metrics matter for accountability: The study evaluated models using multiple metrics (accuracy, precision, recall, F1-score, and error measurements), emphasizing that responsible AI deployment requires transparent reporting of how well systems actually perform across different measures of success.