Over the past month, several companies are competing in the development of AI-based software to detect COVID-19. This article will highlight the software that claims to have the potential in detecting the disease within a matter of a few seconds. Compared to AI-based detection, techniques including the serological, the RT-PCR and a PCR-based test can detect within a few minutes to several days. AI-powered analysis can significantly reduce cost and time. Besides, it has the potential to out-compete the primary method of detection-RT-PCR. These companies are using the repositories of chest scans images to generate algorithms for developing the software. Evidence on the use of chest scan to fight COVID-19 emerged as early as February when manuscripts published in the journal Radiology.
1. In February, Alibaba DAMO Academy announced its AI-based technology that could diagnose COVID-19 cases within 20 seconds with the accuracy rate of 96 per cent.
2. University of Maryland School of Medicine deployed its AI-powered diagnostic solution for COVID-19 called RADLogics, which claims to detect with an accuracy rate of 98 per cent. However, the main concern with this technique is the exposure of radiation dose to a patient. Besides, several national and international experts recommend not to use the CT as a first-line screening test.
3. China-based startup Yitu Healthcare announced its AI-based CT images algorithm for COVID-19 diagnosis.
4. South Korean-based Lunit released its cost-free AI-powered online software based on chest X-ray images to help radiologists in managing COVID-19 cases.
5. The Indian government launched an app called “Aarogya Setu” in a fight against COVID-19. The app alerts individuals when exposed to someone with COVID-19.
6. The Tech giants Apple and Google teamed up to create a contact tracing tool built into the operating system, which probably will work on a similar principle as “Aarogya Setu” app.
7. A Dutch-based company, Delft Imaging released its AI tool, CAD4COVID in triaging COVID-19 cases. The software works by screening the X-ray images for the detection.
8. Voice-based diagnosis- A professor from DY Patil Institute of Bio-Technology and Bio-Informatics, Mumbai, India and a university in Rome and claims that the software can detect COVID-19 suspected cases using a smartphone.
9. IIT-Roorkee, India, developed software using several thousands of X-ray images from COVID-19 patients available in the public domain and claims that the software can detect suspected cases in five seconds.
10. Tianjin Medical Univ Cancer Institute and Hospital, China developed a model based on CT images Model to predict COVID-19 with an accuracy rate of 85.2 per cent, ahead of the pathogenic test.The Longer a patient with COVID-19, the more spots are spread over both lungs. These spots play a vital role in developing the pattern to fed into the software to generate a new algorithm. Therefore, the main concern with some of this software is the detection of suspected cases at the early stage and the possibilities to miss-out cases that are positive.