Predicting coronary heart disease with advanced machine learning classifiers for improved cardiovascular risk...

Area of Science:

• Cardiology
• Medical Informatics
• Artificial Intelligence

Background:

• Coronary heart disease (CHD) is a leading global cause of mortality.
• Early prediction of CHD is crucial for effective clinical intervention.
• Machine learning (ML) shows promise in enhancing diagnostic accuracy for heart disease.

Purpose of the Study:

• To develop a comprehensive ML framework for improved CHD prediction.
• To address challenges in feature selection and class imbalance in healthcare datasets.
• To introduce and evaluate a novel hybrid PSO-ANN model for CHD prediction.

Main Methods:

• Feature selection using mutual information (MI).
• Handling class imbalance with Synthetic Minority Oversampling Technique (SMOTE).
• Developing a hybrid Particle Swarm Optimization-Artificial Neural Network (PSO-ANN) model.

Main Results:

• The proposed PSO-ANN model achieved a prediction accuracy of up to 97%.
• This surpasses the 95.8% accuracy of traditional classifiers like Logistic Regression and Random Forest.
• The framework effectively improved feature selection and addressed data imbalance.

Conclusions:

• The developed ML framework offers superior performance for CHD prediction.
• The hybrid PSO-ANN model demonstrates significant potential for clinical data analysis.
• This approach enhances early detection and decision-making in cardiovascular health.