Prediction of Heart Disease Using Deep Learning and Internet of Medical Things

Abstract

The Internet of Medical Things (IoMT) devices have changed healthcare by providing continuous monitoring of patient physical data. In this case, the prompt and accurate diagnosis of cardiovascular diseases with the aid of focused training programmes has a great potential to enhance patient care. A thorough abstract of a ground-breaking work that predicts heart illness using deep learning and IoMT is presented in this article. In this study is concentrated on the creation and application of a cutting-edge deep learning framework especially created for the IoMT ecosystem's capacity for heart disease prediction. The suggested framework employs convolutional neural networks (CNNs) and recurrent neural networks (RNNs) to the fullest extent possible to extract complex temporal dependencies from the physically heterogeneous data collected by IoMT devices. The biggest accomplishments was the creation of CNN-RNN's new hybrid architecture. This architecture has the ability to extract spatial and sequential characteristics from a variety of patient data flow. To enhance model generalisation, data from several IoMT sources, including pulse oximeters, electrocardiograms, and blood pressure monitors, are seamlessly incorporated. Additionally, the model has improved by the use of transference learning and previously instructed representations from associated medical fields.A large collection of real-world data is used to minutely analyse the proposed model. The results show that it is superior to earlier techniques in terms of anticipated precision and resistance. Additionally, the treatment processes give medical professionals crucial knowledge about the predictive factors that influence the model's judges, which enhances the model's interpretation.