An Intelligent Systems Development for Multi COVID-19 Related Symptoms Detection: A Framework

Thanoon, M. I.

Since it has been announced that COVID-19 is a pandemic, most authorities have been using thermal sensors to detect people with fevers in order to isolate them. Although a wide range of symptoms has been reported in patients confirmed to have COVID-19, fever is the symptom most commonly used for detection in this setting. However, this symptom is common to many other diseases, such as influenza and the common cold. As many studies have suggested, COVID-19 can be detected more accurately if other symptoms are considered. Therefore, the proposed system will concatenate multiple COVID-19-related symptoms for a more accurate classification system. The proposed system will consider the detection of three symptoms: fever, dry cough, and shortness of breath. The proposed system will use a Raspberry Pi and connected sensors to detect these symptoms and then, by using artificial intelligence, will classify people as suspected COVID-19 patients or not. The sensors used to detect the symptoms are a thermal camera, a microphone, an infrared-based camera, and a depth-sensing device. The proposed system was functioning well with accuracies around 97%, 85%, and 96% for fever, cough class, and respiration rate, respectively. © 2021, The Author(s), under exclusive license to Springer Nature Switzerland AG.