Related Experiment Video
Updated: Jan 10, 2026
In Silico Clinical Trials for Cardiovascular Disease
Published on: May 27, 2022
RAVE-HD: A Novel Sequential Deep Learning Approach for Heart Disease Risk Prediction in e-Healthcare
Muhammad Jaffar Khan1, Basit Raza1, Muhammad Faheem2
1Department of Computer Science, COMSATS University Islamabad (CUI), Islamabad 45550, Pakistan.
Insights
RAVE-HD, a novel machine learning approach, enhances heart disease screening by integrating ResNet and Vanilla RNN. It achieves high accuracy and reliability for early detection and clinical decision support.
Area of Science:
- Cardiology
- Artificial Intelligence
- Machine Learning
Background:
- Heart disease (HD) is a leading global cause of mortality, necessitating improved diagnostic tools.
- Existing Internet of Things (IoT)-enabled machine learning for HD screening faces challenges like imbalanced data and feature selection.
- Early and accurate HD diagnosis is crucial for improving patient outcomes.
Purpose of the Study:
- To develop a robust and explainable machine learning approach for enhanced heart disease screening.
- To address limitations in existing HD screening methods, including data imbalance and feature identification.
- To present the RAVE-HD (ResNet And Vanilla RNN Ensemble for HD) approach for improved diagnostic accuracy.
Main Methods:
- A sequential hybrid approach integrating Residual Network (ResNet) and Vanilla Recurrent Neural Network (RNN).
- Preprocessing included duplicate removal, feature scaling, Recursive Feature Elimination, and synthetic data sampling for class imbalance.
- The RAVE model was trained and validated on the HDHI medical dataset, with cross-dataset validation on the CVD dataset.
Main Results:
- The RAVE model achieved 92.06% accuracy and 97.12% ROC-AUC, outperforming baseline models.
- Robustness was confirmed by 10-fold cross-validation, Sensitivity-to-Prevalence analysis, and bootstrap/DeLong tests (p<0.001).
- SHAP analysis provided model interpretability, and cross-dataset validation showed strong generalization (92.4% accuracy).
Conclusions:
- RAVE-HD demonstrates significant promise as a reliable, explainable, and scalable solution for large-scale heart disease screening.
- The approach offers clinically meaningful improvements and acts as a practical decision-support tool for predictive screening.
- RAVE-HD's consistent performance across diverse evaluations and datasets highlights its potential in clinical settings.
Abstract:
Background/Objectives: Heart disease (HD) is recently becoming the foremost cause of death worldwide, underlining the importance of early and correct diagnosis to improve patient outcomes. Although Internet of Things (IoT)-enabled machine learning approaches have demonstrated encouraging outcomes in screening, existing approaches often face challenges such as imbalanced dataset handling, influential feature selection identification, and the ability to adapt to evolving HD data forms. To tackle the aforementioned challenges, we present a sequential hybrid approach, RAVE-HD (ResNet And Vanilla RNN Ensemble for HD), that combines a number of cutting-edge techniques to enhance screening. Methods: Preprocessing phase includes duplicates removal and feature scaling for data consistency. Recursive Feature Elimination is employed to extract the most informative features, while a proximity-weighted random synthetic sampling technique addresses class imbalance to reduce class biases. The proposed RAVE model in RAVE-HD approach sequentially integrates a Residual Network (ResNet) for high-level feature extraction and Vanilla Recurrent Neural Network to capture the non-linearity of the feature relationships present in the HDHI medical dataset. Results: Compared to ResNet and Vanilla RNN baselines, the proposed RAVE model attained superior results: 92.06% accuracy and 97.12% ROC-AUC. Stratified 10-fold cross-validation validated the robustness of RAVE, while Sensitivity-to-Prevalence analysis demonstrated stable recall and predictable precision across varying disease prevalence levels. Additional evaluations, including bootstrap and DeLong analyses, showed statistical significance (p<0.001) of the discriminative gains of RAVE. Minimum Clinically Important Difference (MCID) evaluation confirmed clinically meaningful improvements (≥3%) over strong baselines. Cross-dataset validation using the CVD dataset verified robust generalization (92.4% accuracy). SHAP analysis provided interpretability to build clinical trust. Conclusions: RAVE-HD shows promise as a reliable, explainable, and scalable solution for large-scale HD screening, consistently performing well across diverse evaluations and datasets. Through statistical validation, the RAVE-HD approach emerges as a practical decision-support tool in HD predictive screening results.
