Telemedicine for silent hypoxia: Improving the reliability and accuracy of Max30100-based system

Abstract

This study aims to design, assess, and improve the reliability of the telemedicine-based system for examination and monitoring of the symptoms of silent hypoxia–an extraordinary symptom of COVID-19. We design a telemedicine system for heart rate and oxygen saturation measurement which consists of a photoplethysmograph Max30100 sensor, NodeMCU microcontroller, real-time clock module, firebase realtime database, and Android-based mobile application. The designed system is tested through a comparative test with a commercially available oximeter. A total of 85 experiments from 40 participants in two different positions were conducted. Our analysis shows the accuracy rate of the Max30100 measurement is 97.11% and 98.84%, for heart rate and oxygen saturation (SpO₂), respectively. Bland Altman was used to appraising and visualizing the agreement between the two measurement devices. We further apply calibration to improve the accuracy of the collected data through linear regression, which reveals 97.14% and 99% accuracy data for heart rate and SpO₂, respectively. Finally, a series of end-to-end remote testing is successfully conducted representing the real-life scenario of the telemedicine system. Overall, the designed system attains a reliable option for a telemedicine-based system for examination of the symptoms of silent hypoxia.