Acentric artificial intelligence with deep feature engineering for early heart disease risk prediction

Area of Science:

• Cardiology
• Artificial Intelligence
• Machine Learning

Background:

• Early heart disease identification is crucial for reducing mortality and improving patient outcomes.
• Existing machine learning models for cardiovascular risk prediction often use limited features and fixed rules, hindering adaptability and early detection.
• There is a need for advanced AI frameworks capable of learning complex patterns from clinical data for more accurate risk assessment.

Purpose of the Study:

• To develop and evaluate a novel heart disease prediction framework integrating acentric artificial intelligence (Ac-AI) with deep feature engineering.
• To enhance the accuracy and adaptability of cardiovascular risk prediction models.
• To improve the early detection of heart disease through advanced machine learning techniques.

Main Methods:

• Utilized an autoencoder-based representation learning module for compact latent feature extraction from clinical data.
• Combined learned latent features with original variables to create a comprehensive input set for the Ac-AI classifier.
• Implemented cost-sensitive learning and an adaptive decision threshold within the Ac-AI classifier to boost sensitivity for disease classes.

Main Results:

• The proposed Ac-AI framework demonstrated performance comparable to established models like Random Forest, SVM, and XGBoost on a benchmark heart disease dataset.
• In several experimental runs, the Ac-AI model outperformed baseline machine learning algorithms.
• Consistent results across repeated independent trials indicated the reliability and robustness of the proposed method.

Conclusions:

• Acentric artificial intelligence (Ac-AI) combined with deep feature engineering offers a powerful decision support framework for early heart disease risk prediction.
• The developed framework shows potential for improving the accuracy and early detection capabilities of cardiovascular risk assessment tools.
• This approach represents a significant advancement in leveraging AI for proactive cardiovascular health management.