Data-driven modeling and analysis for cardiovascular disease risk prediction and reduction
Abstract
In recent decades, cardiovascular disease (CVD) has become the leading cause of death in most countries of the world. Since many types of CVD could be preventable by modifying lifestyle behaviors, the objective of this study is to develop an effective personalized lifestyle recommendation approach for reducing the risk of common types of CVD. However, in practice, the underlying causal relationship between the risk factors (e.g., lifestyles, blood pressure, etc.) and disease onset is highly complex. Furthermore, it is also challenging to identify the most effective modification for different individuals by considering both individual’s preferences and the uncertainties in disease progression. Therefore, to address these challenges, this study developed a novel data-driven approach for personalized lifestyle behaviors recommendation based on machine learning and utility function model. The contribution of this work can be summarized into three aspects: (1) a classification-based prediction model is implemented to accurately predict the CVD risk based on the condition of risk factors; (2) a GAN-based approach is developed to capture the relationship between risk factors and generate feasible healthier lifestyle modifications; and (3) a novel personalized evaluation model incorporating utility function is proposed to identify the optimal modification with the consideration of individual’s cost of change and disease progression uncertainty. The effectiveness of the proposed method is validated through an open-access CVD dataset. The results demonstrate that the personalized lifestyle recommended by the proposed methodology can significantly reduce the potential CVD risk. Thus, it is very promising to be further applied to real-world cases for CVD prevention.
Collections
- OSU Theses [15752]