Research on Risk Prediction of Cardiovascular Disease Based on Machine Learning

Cardiovascular disease is one of the leading causes of death worldwide, and its morbidity and mortality are at a high level. Accurately predicting the morbidity risk of cardiovascular disease is of great significance for early intervention and prevention. Traditional diagnostic methods, such as angiography, have high diagnostic accuracy, but they are costly and traumatic. The methods based on electrocardiogram or scoring index depend on the experience of professionals and are difficult to be widely promoted. Over the past few years, the extensive use of electronic health record (EHR) data combined with advancements in machine learning has enabled the prediction of cardiovascular disease risk through deep learning algorithms. In this research, we developed a cardiovascular disease risk prediction model utilizing Long Short-Term Memory Network (LSTM), Convolutional Neural Network (CNN), and Recurrent Neural Network (RNN), and evaluated their comparative performance. Our analysis of EHR data from a major tertiary hospital revealed that the LSTM model achieved the highest precision, recall, and F1 score, with precision at $87.3 \%$, recall at $84.7 \%$, and F1 score at $89.2 \%$. This study offers an effective and accurate machine learning approach for predicting cardiovascular disease risk and provides a theoretical foundation for the development of clinical decision support systems.