SMART DEVICES FOR MONITORING HEALTH AND DISEASES USING MACHINE LEARNING AND DEEP LEARNING METHODS

Abstract

: Worldwide, heart disease is the number one killer. Prediction of heart disease is a very involved process. Early detection of cardiovascular disease signs is one of the most challenging challenges for doctors. Accurate medical choices may be made with the use of heart disease prediction data. Clinical decision-support systems have made extensive use of AI methods for illness prediction and diagnosis. Because of their potential to reveal previously unseen patterns and correlations in medical data given by medical practitioners, these methods are particularly beneficial for creating clinical support systems. A very accurate model is necessary to reduce death rates. Internet of Things (IoT), cloud computing, machine learning, and deep learning methods are employed to construct such precise models. Heart disease symptoms may be reduced and identified with the use of machine learning. The medical decision-making system is meant to aid doctors in their day-to-day work; hence it is an ever-present and regular activity. The accuracy of medical diagnoses is enhanced by the use of web-based healthcare systems. In order to foresee potential issues with clinical risk factors, doctors use a predictive modeling procedure. Early on, a number of different forms of learning technology are used to aid medical professionals in the identification of illness. An accurate, dependable, and continuous monitoring system is also required for prompt intervention and therapy. More strategies need to be developed for prediction models as illness prediction research continues to progress. Decision-making under uncertainty is handled by these models. The study's overarching goal is to develop better methods of early prediction and intervention for cardiovascular illnesses.

We designed a system that uses IoT to monitor patients remotely and accurately assess the degree of their cardiac condition as part of an automated e-healthcare monitoring system. IoT medical sensors capture data on a variety of patient clinical indicators, which are then used in classification algorithms to determine the severity of conditions including hypertension, hypercholesterolemia, and cardiovascular disease. In order to determine the precision with which cardiac illness may be predicted, researchers use machine learning and auto-encoder-based neural network algorithms. Using a total of 35 different clinical criteria, we develop a neutrosophic clinical decision-making system capable of scoring the severity of cardiac disease on a scale from 1 to 5. Furthermore, our suggested models are more accurate and efficient in predicting the existence of heart disease when compared to numerous previous studies.