How Artificial Intelligence Can Help the Global Pandemic

22 October 2020 - Ci Magazine
How Artificial Intelligent Can Help the Global Pandemic
image_pdfimage_print

New Study shows how Artificial Intelligence (AI) can help with the global pandemic. Recently, an (AI) algorithm was able to distinguish between cases of COVID-19, influence, pneumonia, and healthy subjects on CT exams. Chinese researchers published an article online on October 9th yielding a very high accuracy rate.

Not only did it have a very high accuracy rate, but it outperformed five experienced radiologists who participated in the study. The algorithm also showed good generalizability when applied to external test sets, according to first co-authors Dr. Yukun Cao of Tongji Medical College in Wuhan and Cheng Jin of Tsinghua University in Beijing.

The authors write, that since radiologists can perform an individualized diagnosis of COVID-19 with the AI system, this system, could be a new driving force for fighting the global spread of the outbreak.

How did they train and test this system? They used 4,260 CT scans gathered from 3,177 subjects from three centers in Wuhan. Of these studies:

– 2,529 were COVID-19 scans

– 1,338 were cases of community-acquired pneumonia

– 135 were influenza A/B studies

– 258 were normal patients

Another plus is that the AI system is that much faster. It took an average of 2.73 seconds to analyze each study, compared with an average of 6.5 minutes by the radiologists. Although the algorithm performed slightly worse in distinguishing between pneumonia and nonpneumonia, it outperformed the radiologists for the more challenging tasks of distinguishing between community-acquired pneumonia and COVID-19, as well as between COVID-19 and influenza, according to the group.

The AI system even performed with slightly less errors than the radiologists did. Of the 26 errors made by radiologists in distinguishing between COVID-19 and community-acquired pneumonia, 23 (88.5%) were correctly classified by the AI system. Similarly, 20 (86.9%) of the 23 mistakes made by radiologists in distinguishing between influenza and COVID-19 were correctly categorized by the AI software, the researchers noted.

What do these results mean?

These prove that AAI can be used as an effective reader to provide reference suggestions, independently, the authors wrote. It can also screen out suspicious patients for radiologists to confirm/ Or, it can give possible diagnosis error warnings made by radiologists.

As they continue with their research, applying radiomics analysis to Ai results can potentially lead to discovering new biomarkers for COVID-19, which would make our knowledge during this pandemic far improved.

See the published article in Nature Communications for more about the study.