Interpretable machine learning model for cardiovascular disease risk prediction: a feature decomposition-based study.

BACKGROUND: Cardiovascular disease (CVD) is a leading public health issue worldwide. The key to preventing CVD is the early prediction and identification of CVD risk factors. The aim of this study is to construct and validate CVD prediction models using machine learning (ML). METHODS: This study utilized 11 features from 68,205 CVD respondents in the Kaggle dataset. Experiments were conducted using a feature decomposition-based deep learning model (FDDL) to predict CVD incidence in this dataset. The proposed model was compared with six other machine learning models. Moreover, the SHAP method was employed to interpret the model in this study. RESULTS: The FDDL model demonstrated superior predictive capability, achieving benchmark metrics of 75.52% accuracy, 78.14% precision, 71.68% recall, an F1 score of 0.7522, and an AUC-ROC value of 0.7643. In contrast, the LR model exhibited the weakest predictive ability among the compared methods. SHAP value-based feature importance ranking identified diastolic blood pressure, cholesterol level, systolic blood pressure, and age as the most critical predictors for cardiovascular disease risk assessment in our dataset. CONCLUSION: We have developed an ML model for predicting the risk of CVDs. This model shows potential to assist clinicians in identifying high-risk patients and providing a theoretical basis for personalized preventive healthcare measures.