Lunit
Back to List

What do medical studies say about AI for COVID-19 management?

Feb 18, 2021 — 3 min read

Bokyeong Woo

When we released a special version of our chest x-ray AI solution for COVID-19 in March of 2020, we were hoping that the pandemic would be gone by the end of the year. With COVID-19 variants and slow vaccine rollouts, however, it seems clear that COVID-19 is hardly fleeting anytime soon.




A year into the pandemic, lots of medical professionals have been using Lunit AI on a daily basis, to safely and accurately triage and manage COVID-19 patients. There have also been studies published in major peer-reviewed journals that validate the effectiveness of AI in COVID-19 settings. Take a look at the video below to learn about the key takeaways from the studies as well as how AI works in COVID-19 settings.



Click to watch the animated video about Lunit INSIGHT CXR in COVID-19 settings





What do studies say about AI for COVID-19?


Image for post






Performance of Lunit AI in COVID-19 pneumonia detection


Image for post

The performance of Lunit AI in detecting COVID-19 pneumonia compared to that of thoracic radiologists



According to the study published in PLOS ONE, the overall performance of Lunit AI in detecting COVID-19 pneumonia on chest radiography was as high as that of thoracic radiologists. In particular, Lunit AI outperformed thoracic radiologist in sensitivity.

The study demonstrates that AI-aided chest radiography can be a reliable screening exam to identify COVID-19 patients, facilitating patient screening and isolation decisions and ultimately reducing the workload of medical staff during the pandemic.






Turnaround times of AI-aided chest radiography reports


Image for post

The turnaround times of AI-aided CXR reports compared to those of rRT-PCR results



According to the study published in the Korean Journal of Radiologythe turnaround times of AI-aided chest radiography reports were significantly shorter than those of RT-PCR results, with 51 and 507 minutes, respectively.

The study suggests that AI-aided chest radiography interpretation can support timely decision-making and treatment, minimizing the spread of COVID-19 while waiting for RT-PCR results.






Image for post

Online test page of Lunit INSIGHT CXR for COVID-19 (https://www.lunit.io/covid19)






Visit our website dedicated to COVID-19. Upload a chest x-ray image in DICOM standard to see how accurately and quickly AI can detect COVID-19 infected pneumonia.

Though 2021 is going to be full of its travails, we have every hope that it will be immeasurably better for you, for your family and friends, and all of us. We wish you good health and a happy New Year.

AIArtificial IntelligenceChest X RayCovid 19Lunit

More from Blog