An Intelligent Predictive Framework for Early Diagnosis of Cardiovascular Disease Using Deep Neural Network

Abstract

Cardiovascular disease (CVD) is a major cause of death worldwide. Congenital heart disease, arterial disease, heart failure, rheumatic heart condition, and cerebral disease are some of its most prevalent types. Early disease detection can help us avoid potentially fatal diseases and provide patients with better care than we could in later stages because prevention is always preferable than therapy. Those who are diagnosed may have a very high death rate because they are not accessible at an early stage. A variety of research techniques in the machine learning domains can assist in anticipating CVDs and identifying their behavioral patterns in enormous amount of data in order to solve these issues. The results of these estimates will help doctors make judgments and identify patients early, reducing the likelihood of death. This research covers the creation of an innovative, reliable, effective, and intelligent predictive system for early CVD detection using Deep Neural Network (DNN) model in order to optimize prevention and treatment for CVDs. Its goal is to automatically select significant features and detect heart disease in its earlier stages. The presented model's average accuracy, precision, recall, sensitivity, and F1-score are 99.98%, 98.78%, 97.86%, and 98.56%, respectively. Compared to other existing models, the presented method successfully achieved and maximized classification effectiveness with greater amounts of precision and pinpointing.