Machine learning-based telemedicine framework to prioritize remote patients with multi-chronic diseases for emergency healthcare services

AbstractThe remarkable increase in the number of patients poses challenges to healthcare providers in telemedicine systems, and certain challenges are associated with the difficulties of identifying the most urgent emergencies to save lives. Remote monitoring of patients with multiple chronic diseases (MCDs) is a major issue that must be addressed. This work aims to improve the triage system for remote patients who are far from hospital and use telemedicine system by considering the variation in their chronic diseases (chronic heart, hypertension, hypotension, and diabetes diseases). The proposed framework named machine learning-based remote triage framework in telemedicine (ML-ART) collects the patient ’s data using medical sensors and sources within Internet of medical things (IoMT) environment, transfers the data through gateway to telemedicine servers in the hospital, where machine learning is performed to classify (triages) each patient into one of five categories (normal, cold, sick, urgent , and risk) depending on the medical emergency level of the patients. The simulation results showed that the decision tree (DT) algorithm has the most accurate result, 100%, compared to the relevant algorithm (neural network (NN) 97%, support vector machine (SVM) 91%, and random forest (RF) 97%). Th e performance of (DT) logically matched the medical triage procedure. Moreover, the (ML-ART) outcomes improve the performance of e-triage system for remote patients and pave the way for...
Source: Network Modeling Analysis in Health Informatics and Bioinformatics - Category: Bioinformatics Source Type: research