Will artificial software that decodes X-rays be the next step in the fight against the corona virus?
Researchers from the University of Western Scotland (UWS) have developed an artificial intelligence algorithm that can detect within minutes from an X-ray whether a person is sick with corona, flu or pneumonia.
The software compares the X-ray to a pool of 3,000 additional cases and knows how to identify exactly 98 percent of what the patient is suffering from.
However, it is important to note that software that relies on X-rays can not detect a disease in its infancy but only when it is already becoming more significant.
Only then can initial damage to the patient be detected in the photographs.
"At a time when there is a need for rapid and reliable testing even in countries and places around the world where PCR testing is not available due to lack of budget or other reasons, our algorithm provides a quick and accurate answer that can save lives."
Queue for coronation tests, Photo: AFP
In fact, the software was developed for road accident emergencies across the UK.
The algorithm is able to quickly scan the condition of the person injured in the accident and allow him a quick initial diagnosis that saves time for doctors while making rescue efforts.
However, when it became clear that the antigen tests were not accurate enough and that there was a lack of PCR tests, the researchers decided to use the system for Corona.
Professor Naim Ramzen, director of the university's computer research center, explained: "Now that the epidemic has created a need for a reliable tool that can detect the virus in the body, this is even more true in the current omicron wave, along with many countries unable to perform large corona tests and few laboratories "In this respect, our technology uses cheap and very accessible equipment with accurate results. It is true that this is a technology that does not detect the disease in its beginning but it can certainly save lives by diagnosing severe cases of the virus."
Were we wrong?
Fixed!
If you found an error in the article, we'll be happy for you to share it with us