Smart Health Monitoring System with Deep Learning Models for Automated Heart Disease Detection

Abstract

Heart disease has remained a major cause of mortality globally and therefore there is the important necessity of early diagnosis and follow-up health studies. Conventional diagnostic methods are also time-consuming, costly and require the interpretation of the expert, and thus they are not suitable in mass screening. In this study, a Smart Health Monitoring System is presented to overcome such problems. This is by incorporating data collection which is enabled by Internet of Things (IoT) and a CNN model to conduct automated detection of cardiac illness. Before the model is trained, the Heart Disease Dataset that is composed of thirteen clinical attributes was preprocessed by data cleansing, normalization, feature extraction, and balancing. As an insistence of its resilience and usefulness in diagnosing heart disease, the proposed CNN model showed good prediction accuracy (ACC) of 99.9, 99.5, 98.4, and F1-score (F1) of 99.4. The CNN model was much better than the more traditional models such as Naive Bayes (NB), Random Forest (RF) and Logistic Regression (LR) in all measures of assessment. All things considered, findings demonstrate that the suggested Smart Health Monitoring System based on CNN has great promise as an intelligent, dependable, and real-time answer to the problems of early cardiac risk assessment and preventative healthcare.