New artificial intelligence can detect colorectal cancer in less than a second, researchers say

Brett Molina

Japanese researchers demonstrated an artificial intelligence capable of identifying and analyzing polyps found during a colonoscopy in less than a second.

The computer-assisted diagnostic system was revealed during the United European Gastroenterology Week, running through Wednesday in Barcelona.

The endoscopic system uses a magnified view of a colorectal polyp to study its features and compare it with 30,000 endocytoscopic images used for machine learning.

Researchers said they were able to predict the pathology of the polyp in less than a second, with 86% accuracy, based on a study assessing more than 300 polyps.

"The most remarkable breakthrough with this system is that artificial intelligence enables real-time optical biopsy of colorectal polyps during colonoscopy, regardless of the endoscopists’ skill," said Dr. Yuichi Mori, a researcher from Showa University in Yokohama, Japan, and study lead, in a statement.

Mori said researchers now want to work on a broader study aimed at creating a system that can automatically detect polyps.

Its value to the medical world is a key reason why supporters such as Facebook CEO Mark Zuckerberg are excited about the future of artificial intelligence. During a Facebook Live chat in July, Zuckerberg discussed how AI can create safer cars and diagnose diseases earlier.

"I’m just much more optimistic in general on this," he said.

Critics of AI, most notably Tesla CEO Elon Musk, said governments should regulate how artificial intelligence is built and used to prevent potential global disasters.

Follow Brett Molina on Twitter: @brettmolina23.