What is the clinical evidence from this study?

Study design: Other. Population: Cardiovascular disease (n=918). Intervention: Interpretable and responsible AI framework (Causal-Guided Ensemble Voting Classifier). Primary outcome: Diagnostic accuracy for cardiovascular disease.

What does this research mean for the field?

A hybrid ensemble AI framework integrating causal inference achieves 89% accuracy in predicting cardiovascular disease while identifying actionable predictors such as exercise-induced angina. Novelty: ClaimNovelty.METHODOLOGICAL. Consensus alignment: ConsensusAlignment.SUPPORTS_CONSENSUS.

What question did this study set out to answer?

The aim is to improve the diagnosis of cardiovascular disease using machine learning to facilitate early detection.

April 3, 2026Open Access

Advancing cardiovascular disease diagnosis with an interpretable and responsible AI framework

Key Result

A hybrid ensemble AI model integrating causal inference achieved 89% accuracy in predicting cardiovascular disease while identifying actionable predictors like exercise-induced angina.

Key Points

The aim is to improve the diagnosis of cardiovascular disease using machine learning to facilitate early detection.
Developed a machine learning ecosystem incorporating non-clinical and clinical data.
Utilized advanced methods like tabular neural networks and ensemble algorithms validated on multi-regional datasets.
Incorporated interpretability tools and fairness measures to enhance clinical trustworthiness.
The model achieved 89% accuracy in diagnostics while identifying key predictors like exercise-induced angina.
Shapley analysis revealed ECG features as significant predictors, especially ST-segment slope and ST depression.
Counterfactual analysis indicated that lifestyle changes could shift risk predictions from diseased to healthy.

Structured PICO

Does an interpretable and responsible AI framework improve the accuracy and fairness of cardiovascular disease diagnosis?

Population

918 instances from the UCI Heart Disease dataset, aggregated from four sources (Cleveland, Switzerland, Long Beach VA, and Hungary).

Intervention

Machine learning ecosystem including tabular neural networks and ensemble methods (XGBoost, Random Forest) with interpretability tools (SHAP, counterfactuals), fairness mitigation (FairLearn), and uncertainty quantification.

Outcome

Classification performance (Accuracy, AUC, F1 Score, Precision, Recall) for cardiovascular disease prediction

An interpretable AI framework integrating clinical and non-clinical data achieved high diagnostic accuracy for cardiovascular disease while addressing ethical deployment through fairness mitigation and uncertainty quantification.

Limitations

Reliance on retrospective datasets requiring synthetic data augmentation (SMOTE and CTGAN) to address class imbalances and demographic bias.
Presence of regional biases in the source data requiring unsupervised clustering to generalize the model.

Abstract

Cardiovascular disease (CVD) remains a leading global health threat, responsible for one in five deaths worldwide. Early detection is critical to mitigate morbidity and mortality, yet traditional diagnostic methods often rely on reactive clinical assessments, missing opportunities for preventive intervention. In this study, a machine learning (ML) ecosystem is developed to enhance CVD diagnosis through two key approaches: (1) an early warning system using non-clinical, self-reported features for accessible risk stratification, and (2) specialized diagnostic models integrating clinical and non-clinical data. The framework leverages advanced ML techniques, including tabular neural networks (TabNet, TabPFN) and ensemble methods (XGBoost, Random Forest), validated on multi-regional datasets. Shapley Additive Explanations (SHAP) analysis identified ECG-related features as dominant predictors of CVD risk, with ST-segment slope (+0.93) and ST depression (+0.63) exhibiting the strongest effects. Counterfactual explanations from the non-clinical model further revealed actionable preventive measures: reducing exercise-induced angina and chest pain severity, alongside increasing exercise heart rate, could shift predictions from diseased to healthy, highlighting the model’s utility for lifestyle interventions. To address ethical and clinical trustworthiness, interpretability tools (SHAP, counterfactuals), fairness mitigation (FairLearn), and uncertainty quantification (Bayesian Neural Networks) are incorporated. Causal inference identified key predictors and their Average Treatment Effects (ATEs) such as exercise-induced angina (ATE: 0.36) and ST slope (ATE: 0.33), informing a hybrid ensemble model that achieved 89% accuracy while reducing dimensionality. The system aligns with FDA Good ML Practices and EU Trustworthy AI guidelines, offering a scalable solution for early detection and equitable diagnosis.

Mark Helpful

Bookmark

Relay

View Full Paper