Health fog: A deep learning approach leveraging Internet of Things and fog computing for real-time heart disease prediction

Real-time data available from IoT devices, allows predicting a patient’s risk of heart disease using Health Fog. The proposed approach is economically affordable yet reliable, as it offers low-cost input devices to meet the necessary requirements. The goal achieved is accurate results with low latency. The reduction in latency time was achieved through fog processing. The Arduino UNO serves as an IoT device and the Raspberry Pi serves as a node in the fog architecture. The centralised cloud system has its limitations in terms of direct access, hence the use of a distributed multilayer model significantly improves the availability of computing resources locally where they are needed. Heart disease is predicted using a deep learning algorithm. The DNN algorithm is used by the model to improve performance. The data is sent and received by an Android mobile phone, a fog node that completes the forecast sends it back. Deep learning is accepted for predicting outcomes in the healthcare industry because it is characterised by precise results. Requests are sent between the client and server computers via the REST API. As a result, data is digitally stored in the cloud for later use. The model makes use of Bagging Classifier ensemble learning. Based on a deep learning algorithm, the user submitting input for heart disease prediction gets a result with high reliability.