What is the clinical evidence from this study?

Study design: Observational. Population: Heart disease. Intervention: Uncertainty-Aware Feature-Weighted Ensemble (UAFE) vs. state-of-the-art baselines. Primary outcome: accuracy of heart disease prediction.

What question did this study set out to answer?

The central aim is to enhance heart disease diagnosis by addressing clinical reliability and feature heterogeneity in predictions.

March 15, 2026Scientific Reports0 citationsOpen Access

Uncertainty-aware feature-weighted ensemble framework for heart disease prediction

XWXinling WangChinese PLA General Hospital YFYangrui FanMacau University of Science and Technology MYMinglian YuThe 309th Hospital of Chinese People's Liberation Army

Key Result

UAFE achieved an accuracy of 86.60% in heart disease prediction, outperforming traditional models significantly.

Key Points

The central aim is to enhance heart disease diagnosis by addressing clinical reliability and feature heterogeneity in predictions.
Developed an uncertainty-aware feature-weighted ensemble (UAFE) framework.
Implemented feature stratification for tailored learning on important subgroups.
Employed dynamic uncertainty weighting to modify model contributions based on prediction disagreements.
Utilized neighborhood refinement to correct decisions for uncertain predictions using local geometric information.
Achieved an accuracy of 0.8660 in multi-center experiments.
Demonstrated a mean accuracy of 0.8332 during leave-one-center-out validation across four clinical sites.
Established superior stability against distribution shifts in clinical environments.

Study Design

Type

Observational

Multicenter

Yes

Structured PICO

Does the Uncertainty-Aware Feature-Weighted Ensemble (UAFE) framework improve prediction accuracy and reliability for heart disease compared to traditional machine learning models?

Population

Multi-center dataset from four international clinical sites for heart disease prediction

Intervention

Uncertainty-Aware Feature-Weighted Ensemble (UAFE) framework integrating feature stratification, dynamic uncertainty weighting, and neighborhood refinement

Comparator

Baseline machine learning models (KNN, SVM, Logistic Regression, TabNet, LightGBM, XGBoost, Random Forest, Voting Ensemble)

Outcome

Prediction accuracy

The UAFE machine learning framework enhances heart disease prediction accuracy and reduces false negatives by dynamically adjusting to prediction uncertainty and feature heterogeneity.

Abstract

This paper considers the challenge of clinical reliability and generalization in heart disease diagnosis, where traditional ensembles often overlook feature heterogeneity and lack uncertainty quantification. We propose an uncertainty-aware feature-weighted ensemble (UAFE) framework designed for robust cardiovascular prediction. UAFE integrates three core components: (1) Feature Stratification to train specialized learners on importance-based subgroups; (2) Dynamic uncertainty weighting to adaptively adjust model contributions based on prediction disagreement; and (3) Neighborhood refinement to rectify decisions for high-uncertainty samples using local geometric priors. Comprehensive experiments on a multi-center dataset demonstrate that UAFE outperforms state-of-the-art baselines with an accuracy of 0.8660. Furthermore, leave-one-center-out (LOCO) validation across four international clinical sites shows that UAFE maintains superior stability (mean accuracy 0.8332) against institutional distribution shifts, establishing its effectiveness for deployment in unseen clinical environments.

AI에게 질문

Bookmark

View Full Paper