Researchers train AI to diagnose diseases | Inquirer Technology

Researchers train AI to diagnose diseases

/ 02:21 PM June 21, 2016

In the future, AI may work hand in hand with doctors to diagnose and treat diseases more effectively. Image

AI development is currently in full swing, with scientists teaching computers how to diagnose diseases.

According to a report on Engadget, researchers from Harvard Medical School and Beth Israel Deaconess Medical Center have developed a method for training artificial intelligence based on deep learning.


They started by feeding the AI hundreds of slides that indicated which parts had cancerous cells and which were normal ones. Afterwards they filtered out those that the machine had trouble with and proceeded to feed more difficult samples. In the end, the AI was able to attain 92% accuracy. A trained pathologist, on the other hand, is accurate 96%.

The team was able to prove the effectivity of their methodology during a competition at the International Symposium of Biomedical Imaging. The AI was asked to look for great cancer in lymph nodes.


What excited the researchers the most was when the pathologists’ and the AI results were combined, it yielded a 99.5 percent accuracy. It would seem that having two heads is indeed better than one, at least when spotting cancer in patients. Alfred Bayle

The latest tech news delivered to your inbox

By providing an email address. I agree to the Terms of Use and acknowledge that I have read the Privacy Policy.
TOPICS: AI, Artificial Intelligence, Harvard, Medical research
Read Next
Don't miss out on the latest news and information.

Subscribe to INQUIRER PLUS to get access to The Philippine Daily Inquirer & other 70+ titles, share up to 5 gadgets, listen to the news, download as early as 4am & share articles on social media. Call 896 6000.

For feedback, complaints, or inquiries, contact us.

Subscribe to our technology news

By providing an email address. I agree to the Terms of Use and
acknowledge that I have read the Privacy Policy.

© Copyright 1997-2022 | All Rights Reserved

We use cookies to ensure you get the best experience on our website. By continuing, you are agreeing to our use of cookies. To find out more, please click this link.