A Microcontroller Based Bluetooth Pulse Oximeter with Adaptive Algorithm for Non-Invasive COVID-19 Detection

Abstract

A pulse oximeter is a clip-on device that is used to measure the oxygen saturation of the arterial blood. This measurement of oxygen  saturation is very important to the medical practitioners because there are many anomalies that can lead to a drop of oxygen saturation  such as suffocation, choking, lung infections (ronchitis, pneumonia and bronchiolitis), COVID-19 and inhalation of poisonous chemicals.  With the outbreak of the novel Covid-19 pandemic which has in turn led to a paradigm shift in the way that activities are being carried out  in different economies of the world, it has indeed become an imperative to adopt a more pragmatic approach in not only combating  the thread of the current pandemic but to also adequately prepare ahead for any other outbreak that may occur in the future. A lot of  researchers have adopted the use of traditional pulse oximeter for getting the level of oxygen saturation of the blood via non-invasive  means. However, the use of traditional pulse oximeter exposed the medical practitioners to infectious diseases due to proximity problem.  Also, discrepancies in the environmental factors, the readings obtained from the device becomes flawed with errors. To solve  the aforementioned problem, this work designed a robust wireless pulse oximeter which is a very important device used in the medical  field to obtain the level of oxygen saturation and heartbeat of a patient with safe proximity. Performance evaluation of the designed  pulse oximeter was carried out using the SpO₂ and heart rate as performance metrics. It can be observed that the Root Mean Squire  Error (RMSE) values of SpO₂ and BPM were found to be 1.549 and 3.233 respectively. The small RMSE values obtained indicates the  successful performance of the designed pulse oximeter.