Arman Process Journal (APJ)

Abstract One of the most important applications of the Internet of Things in the field of health is remote monitoring of patients. This technology allows doctors to check the health status of patients in real time, which is especially vital for people suffering from or prone to heart diseases. Prediction of cardiovascular diseases is known to be a complex challenge that faces low accuracy in existing models. In this research, a new recommender system for predicting cardiovascular diseases is proposed that uses a convolutional neural network to analyze physiological data of patients. Physiological data from patients are collected remotely through four biological sensors including ECG sensor, blood pressure sensor, heart rate sensor and blood sugar sensor. These data are then processed by an Arduino controller and the convolutional neural network model is used to predict cardiovascular disease. With outstanding capabilities in extracting local features and without the need for complex time sequence analysis, this model can effectively use fixed numerical data such as blood pressure, heart rate, and blood sugar to diagnose heart diseases. The experimental results showed that the convolutional neural network was able to effectively extract local and non-temporal features of the data and help the model achieve a prediction accuracy of 98.90%.