Machine learning techniques for improved prediction of cardiovascular diseases using integrated healthcare data

Area of Science:

• Medical Informatics
• Machine Learning in Healthcare
• Cardiovascular Health

Background:

• Cardiovascular disease (CVD) poses a significant global health burden.
• Early detection of CVD is crucial for effective management and improved patient outcomes.
• There is a growing need for advanced, reliable diagnostic tools to supplement traditional methods.

Purpose of the Study:

• To explore the potential of machine learning (ML) algorithms for precise cardiovascular disease diagnosis.
• To develop and evaluate ML models using integrated, large-scale healthcare datasets.
• To identify the most effective ML model for predicting cardiovascular disease risk.

Main Methods:

• Collected and integrated extensive, publicly available healthcare records from multiple databases.
• Performed data preprocessing including cleaning, feature alignment, and handling missing values.
• Trained and evaluated several machine learning models on a dataset of 311,710 samples.

Main Results:

• The CatBoost machine learning model achieved the highest performance in predicting cardiovascular disease.
• The CatBoost model obtained an Area Under the Curve (AUC) of 94.1%.
• This indicates a high level of accuracy and robustness in the model's diagnostic predictions.

Conclusions:

• Machine learning decision support systems show significant promise for cardiovascular disease diagnosis.
• The CatBoost model offers a robust and accurate approach for early CVD detection.
• Leveraging large healthcare datasets with ML can enhance diagnostic precision and patient care.