Evaluation of coronary heart disease risk prediction based on simple physical examination parameters by machine learning model: a retrospective cohort model development and validation study - Report - MDSpire
Advertisement
Evaluation of coronary heart disease risk prediction based on simple physical examination parameters by machine learning model: a retrospective cohort model development and validation study
Clinical Report: Development and Validation of a Machine Learning Model for Predicting Coronary Heart Disease Risk
Overview
This study developed and validated a machine learning model for predicting coronary heart disease (CHD) risk using routine clinical indicators. The model demonstrated strong performance in both internal and external validation cohorts, highlighting its potential as a clinically interpretable tool for cardiovascular risk assessment.
Background
Coronary heart disease (CHD) is a leading cause of mortality globally, with significant healthcare implications. Traditional risk assessment tools often fail to capture complex interactions among risk factors, leading to delayed diagnoses and suboptimal management. Machine learning offers a promising approach to enhance predictive accuracy and clinical applicability in CHD risk stratification.
Data Highlights
Validation Type
AUC
Accuracy
F1 Score
Internal
0.977
0.942
0.944
External
0.929
0.885
N/A
Key Findings
The model achieved an AUC of 0.977 in internal validation and 0.929 in external validation.
Accuracy rates were 94.2% for internal validation and 88.5% for external validation.
Key predictors identified included systolic blood pressure, age, total cholesterol, and fasting glucose.
SHAP analysis revealed nonlinear effects and interactions among predictors.
The model utilized a stacked ensemble approach combining gradient boosting, random forest, and XGBoost.
Clinical Implications
The validated machine learning model provides a robust tool for clinicians to assess CHD risk using readily available clinical metrics. Its strong performance in diverse cohorts suggests it could enhance early detection and management strategies for cardiovascular disease.
Conclusion
This study presents a machine learning model that effectively predicts CHD risk, demonstrating both high accuracy and generalizability. Such models could significantly improve cardiovascular risk assessment in clinical practice.
