Review of Method to Automatic Detection of COVID-19

Leyang, L.; Guixing, C.; Jun, L.

COVID-19 has begun to spread around the world, and this situation is so serious that we need to find the patient as soon as possible. To detect COVID-19 at an early stage, artificial intelligence (AI) is applied because of its powerful data processing capabilities. Therefore, many scholars have designed algorithms to automatically identify whether they are patients with COVID-19 from CT or X-ray of the lungs. Although nucleic acid detection is the main method for the diagnosis of COVID-19, lung image detection is still a method worthy of consideration, and X-ray would still be useful in the evaluation of therapy for the patients who are confirmed and hospitalized. This paper examines the state-of-the-art method for the automatic diagnosis of COVID-19 and provides a detailed description of its model and results, and expresses our opinions on it. Then, we list some common CT and X-ray databases. Most importantly, we conducted a survey on issues related to the automatic detection of COVID-19 for some people who are computer vision professionals and have published the results of the survey. This paper is an all-around review of the automatic detection methods and related survey of COVID-19, which will play an enlightening role in future research. © 2021 ACM.