Machine Learning-Based Model for Predicting Coronary Heart Disease Using Preβ HDL and Cytokines as Plasma Biomarkers.

Area of Science:

• Cardiovascular Disease Research
• Biomarker Discovery
• Machine Learning in Medicine

Background:

• Coronary heart disease (CHD) is a leading global cause of mortality.
• Improved prediction and early intervention methods for CHD are crucial.
• Plasma biomarkers, including Preβ High-Density Lipoprotein (HDL) and cytokines, are being investigated for CHD risk assessment.

Purpose of the Study:

• To evaluate the predictive efficacy of plasma Preβ HDL and cytokines for CHD.
• To utilize machine learning (ML) algorithms for enhanced CHD risk prediction.
• To identify specific biomarkers that can discriminate between CHD patients and control subjects.

Main Methods:

• A case-control study involving 108 individuals (41 CHD, 67 controls).
• Analysis of 35 plasma cytokines and Preβ HDL levels.
• Application of random forest algorithms with feature engineering and synthetic data augmentation to 20,000 samples.

Main Results:

• Individuals with CHD had significantly higher plasma Preβ HDL levels compared to controls.
• A random forest model incorporating Preβ HDL and specific cytokines (FGF-Basic, MCP-1, Eotaxin, IL-10, IL-9, IL-1β) achieved 100% F1 score, accuracy, and AUROC.
• These findings highlight the strong predictive power of these biomarkers.

Conclusions:

• Plasma Preβ HDL and specific cytokines are highly effective predictors of CHD when analyzed with ML.
• Further research into these biomarkers could identify novel drug targets for CHD.
• ML methodologies enhance the potential of biomarker-based CHD risk assessment.