Invasive and non-invasive variables prediction models for cardiovascular disease-specific mortality between machine
Seonggyu Choi1, Minsuk Oh1,2,3, Dong Hoon Lee1
1Department of Sports Industry Studies, Yonsei University, Seoul, South Korea.
Insights
Predicting cardiovascular disease (CVD) mortality is possible using non-invasive indicators alone. Machine learning models offer slightly better prediction than traditional methods, without needing blood lipid profiles.
Area of Science:
- Cardiology
- Biostatistics
- Machine Learning in Healthcare
Background:
- Cardiovascular disease (CVD) remains a leading cause of mortality globally.
- Accurate prediction of CVD-specific mortality is crucial for timely intervention.
- Traditional statistical models often rely on invasive biomarkers, necessitating further research into non-invasive predictors.
Purpose of the Study:
- To evaluate the predictive performance of cardiovascular disease (CVD)-specific mortality using non-invasive indicators.
- To compare traditional statistical models with machine learning approaches for CVD mortality prediction.
- To determine if incorporating blood lipid profiles significantly enhances predictive accuracy.
Main Methods:
- Utilized data from 1,749,444 Korean adults with a 10-year follow-up for CVD-specific mortality.
- Employed traditional Cox proportional hazards models and machine learning models (Random Survival Forest, Gradient Boosting Survival, Survival Tree).
- Compared model performance using Area Under the Curve (AUC), c-index, and Brier score, with and without invasive variables (triglycerides, fasting glucose, cholesterol).
Main Results:
- All models using only non-invasive predictors (sex, age, waist-to-height ratio, diabetes, hypertension, physical activity) achieved AUCs > 0.800.
- Non-invasive models demonstrated performance comparable to models including blood lipid profiles.
- Machine learning models exhibited marginally superior predictive performance over time compared to traditional models, though differences were not substantial.
Conclusions:
- Non-invasive indicators are sufficient for valid prediction of CVD-specific mortality.
- Machine learning models provide a slight, but not substantial, improvement in prediction accuracy.
- The addition of invasive blood lipid profiles does not significantly enhance the predictive performance of CVD mortality models.
Abstract:
This study examined the predictive performance of cardiovascular disease (CVD)-specific mortality using traditional statistical and machine learning models with non-invasive indicators, and assessed whether adding blood lipid profiles improves prediction. Data were from 1,749,444 Korean adults (44.7% female) from the Korea Medical Institute. Non-invasive predictors included sex, age, waist-to-height ratio, diabetes, hypertension, and physical activity; invasive variables included triglycerides, fasting glucose, and cholesterol. CVD-specific mortality was tracked over a 10-year follow-up. We applied Cox proportional hazards models (with and without elastic net penalty), Random Survival Forest, Gradient Boosting Survival, and Survival Tree models. Predictive performance was compared using area under the curve (AUC), c-index, and Brier score. All models using only non-invasive predictors achieved AUCs > 0.800 and were not inferior to models including blood profiles. Machine learning models showed slightly higher predictive performance over time than traditional models, but differences were not substantial. Both approaches appear valid for predicting CVD-specific mortality using non-invasive data. Machine learning models may offer marginally improved prediction, and the addition of invasive variables may not substantially enhance model performance.
Related Concept Videos
01:25Statistical Methods for Analyzing Epidemiological Data
01:20Comparing the Survival Analysis of Two or More Groups
01:18Steps in Outbreak Investigation
01:23Mechanistic Models: Compartment Models in Individual and Population Analysis
01:26Pharmacokinetic Models: Comparison and Selection Criterion
Physiological models take a detailed approach by considering specific molecular processes. They can predict drug distribution, metabolism, and elimination changes, providing a comprehensive understanding of how drugs interact with the body.