Data-Mining-Based Coronary Heart Disease Risk Prediction Model Using Fuzzy Logic and Decision Tree

Area of Science:

• Cardiovascular Health
• Artificial Intelligence in Medicine
• Public Health Informatics

Background:

• Coronary heart disease (CHD) prediction is crucial in Korea, yet research is limited.
• A need exists for effective CHD prediction and classification methods tailored for the Korean population.

Purpose of the Study:

• To develop a rule-based coronary heart disease (CHD) prediction model for Koreans.
• To enhance CHD prediction accuracy by integrating fuzzy logic and decision tree techniques.

Main Methods:

• A decision tree (Classification and Regression Tree [CART]) approach was employed to generate predictive rules.
• Fuzzy logic was incorporated to address uncertainties inherent in coronary heart disease (CHD) prediction.
• The model was trained using data from the Korean National Health and Nutrition Examination Survey VI (KNHANES-VI).

Main Results:

• The developed model achieved an accuracy of 69.51%.
• The receiver operating characteristic (ROC) curve value was recorded at 0.594.
• The proposed fuzzy logic and CART-driven model demonstrated superior efficiency over alternative models.

Conclusions:

• The integrated fuzzy logic and decision tree model offers a more efficient approach to coronary heart disease (CHD) prediction in Koreans.
• The findings highlight the potential of AI-driven methods for improving cardiovascular disease risk assessment.