DISTRIBUTED INTEROPERABILITY SOLUTIONS SMART HEALTHCARE SYSTEM FOR MULTI-PATIENT VITAL SIGNS MONITORING AND FORECAST OF CRITICAL ALERTS

Abstract

In current healthcare systems, doctors and paramedics have to individually observe the readings on attached devices to judge a patient’s condition. A smart healthcare system may generate an opinion about the patient by compiling data from various vital sign monitoring devices. However, multi-vendor devices face the data interoperability problem because of the varying standards
used. Current solutions rely on centralized cloud/fog-based servers for interoperability which is a barrier to real-time multi-patient monitoring. This research presents an Edge-computing based distributed interoperability framework for smart healthcare devices and presents a system that continuously monitors the patients’ vital signs, ensemble the results for display in the nursing office, or in the doctor’s wallet. A healthcare setup was emulated for testing the proposed solution. Results are compared to the centralized authority-based system, which showed that the proposed solution performed better in terms of response time, with the advantage of utilizing the local resources to achieve data interoperability. We used deep learning techniques to learn patients’ critical situation from the vital signs monitoring database and predicted the critical situations to alert about the critical
patients with 86% accuracy and 91% precision. Our model achieved a sensitivity of 90% and specificity of 72%. Hence a good overall performance has been achieved.