An Intelligent System for Cardiovascular Disease Prediction Using Deep Learning

Abstract

Heart disease remains among the major causes of death in the world. Hence, accurate and reliable prediction schemes are key. To detect the presence of diseases in early stages, data mining and deep learning algorithms offer powerful tools to determine important trends in large and complex data. To enhance the prediction of cardiac diseases, this paper proposes a hybrid deep learning model, which is the combination of Convolutional Neural Networks (CNN) and Long Short-Term Memory (LSTM) networks. The model is implemented by using Python, and its effectiveness is evaluated by such performance measures as recall, accuracy, and precision. According to experimental data, the hybrid CNN–LSTM model outperforms a standalone CNN by 8%, achieving 98% training accuracy and 97% validation accuracy. Additionally, the ROC curve shows an improvement in true positive rate of 0.83, and the model achieves precision and recall values of 0.97 each. These results show that the suggested hybrid strategy outperforms traditional deep learning models in terms of predictive power, facilitating better clinical judgment in the prognosis of cardiovascular disease.