[ad_1]
Northwestern University (NU) of America researchers have developed an electronic platform using the “deep learning” artificial intelligence program and named it “DeepCovid-XR”, so that it can
Researchers at Northwestern University (NU) of America developed an electronic platform using artificial intelligence’s “deep learning” program, which they called “DeepCovid-XR,” so that it could detect the Corona virus by analyzing images at X-rays of patients ‘lungs and more Physicians’ accuracy and speed.
According to the news site “The Express Tribune”, the experiments showed that the machine learning algorithm surpassed the team of chest radiologists by detecting the emerging corona virus in chest X-ray images, about 10 times faster and by 6% more accurate than doctors. .
The researchers used 17,002 chest X-ray images to develop, train and test the new algorithm, and of those images, 5,445 patients tested positive for the coronavirus.
Researchers compared the platform’s performance to the performance of five radiologists and cardiothoracologists, analyzing 300 random patient images from Lake Forest Hospital in the United States.
It took each radiologist two to three and a half hours to examine this set of images, while the AI system took only about 18 minutes.
The radiologist’s accuracy was between 76 and 81%, while the “DeepCovidD-XR” platform performed slightly better with 82% accuracy.
“X-rays are inexpensive and are already a common component of routine care, as the artificial intelligence platform can save money and time, because timing is very important when dealing with Corona virus in patients.” , said Agilios Katsagilos, an artificial intelligence expert and lead author of the study.
The “DeepCovid-XR” platform is still under development, but it is likely that it will soon be used in human clinical trials.
.
[ad_2]
Source link
