Enhancing clinically cardiovascular machine learning model for risk prediction via sample augmentation
By
Xiaoyu Tang
Min Tang
Wu Liu
Shaoyang Cui
June 9, 2026
Clinical Scorecard: Improving Cardiovascular Risk Prediction through Machine Learning by Utilizing Sample Augmentation Techniques
At a Glance
Category Detail
Condition Cardiovascular Disease Key Mechanisms Machine learning models utilizing sample augmentation techniques for risk prediction. Target Population Individuals at risk for cardiovascular disease. Care Setting Clinical settings utilizing structured clinical data.
Key Highlights
Moderate data augmentation (2×) improves model robustness and accuracy. Random Forest (RF) achieved 94.0% accuracy and 95.9% sensitivity. SHAP and PDP methods provide interpretable risk factor profiles. Guideline-Based Recommendations
Diagnosis
Utilize machine learning models for early risk identification. Management
Implement interpretable risk models for clinical decision support. Monitoring & Follow-up
Evaluate model performance using MAE, RMSE, and R2 metrics. Risks
Address high variance and instability due to small sample sizes. Patient & Prescribing Data
Patients with cardiovascular risk factors.
Machine learning can enhance risk stratification and treatment response assessment.
Clinical Best Practices
Employ feature-space constrained sample augmentation for model training. Combine SHAP and PDP for comprehensive interpretability of risk factors. Related Resources & Content
