PREDICTING CARDIOVASCULAR DISEASE RISK USING ARTIFICIAL INTELLIGENCE: A HYBRID ENSEMBLE MODEL AND EXPLAINABILITY-BASED APPROACH
Received: 2026-05-23
Published: 2026-06-06
Abstract
This article presents a hybrid artificial-intelligence model that integrates tabular clinical features (XGBoost), electrocardiographic time series (BiLSTM with attention) and medical imaging (CNN) for early and accurate prediction of cardiovascular disease (CVD) risk. The model was trained on a combined cohort drawn from MIMIC-IV, UK Biobank and PTB-XL (test set n = 10,832) and complemented by SHAP-based explainability and a Federated Learning paradigm. The hybrid model achieved AUC-ROC = 0.968, sensitivity 92.7% and Brier score = 0.049, outperforming the conventional Framingham Risk Score across all metrics (p < 0.001). The findings demonstrate the clinical potential of a multimodal, interpretable and privacy-preserving approach.
Keywords
List of references
-
Roth, G. A., Mensah, G. A., Johnson, C. O. et al. (2020). Global burden of cardiovascular diseases and risk factors, 1990–2019. Journal of the American College of Cardiology, 76(25), 2982–3021.
-
Krittanawong, C., Virk, H. U. H., Bangalore, S. et al. (2020). Machine learning prediction in cardiovascular diseases: a meta-analysis. Scientific Reports, 10(1), 16057.
-
Shah, P., Shukla, M., Dholakia, N. H., & Gupta, H. (2025). Predicting cardiovascular risk with hybrid ensemble learning and explainable AI. Scientific Reports, 15(1), 17927.
-
Lundberg, S. M., & Lee, S. I. (2017). A unified approach to interpreting model predictions. Advances in Neural Information Processing Systems, 30, 4765–4774.
-
McMahan, H. B., Moore, E., Ramage, D., Hampson, S., & Agüera y Arcas, B. (2017). Communication-efficient learning of deep networks from decentralized data. AISTATS, 54, 1273–1282.
-
Dwork, C., McSherry, F., Nissim, K., & Smith, A. (2006). Calibrating noise to sensitivity in private data analysis. TCC 2006, LNCS 3876, 265–284.
-
Hannun, A. Y., Rajpurkar, P., Haghpanahi, M. et al. (2019). Cardiologist-level arrhythmia detection using a deep neural network. Nature Medicine, 25(1), 65–69.
-
DeLong, E. R., DeLong, D. M., & Clarke-Pearson, D. L. (1988). Comparing the areas under two or more correlated receiver operating characteristic curves. Biometrics, 44(3), 837–845.
About the Authors
License
This work is licensed under a Creative Commons Attribution 4.0 International License.