AI detecting more deadly polyps than ever before

SALT LAKE CITY, Ut. (WALB) - Artificial Intelligence (AI) is transforming the field of medicine becoming part of everything from diagnosis to treatment to research.

It gathers vast amounts of medical data to identify patterns, such as which cancers will respond to which treatments better, or which hip replacement will work best for your body type. And now, AI is working to make colonoscopies better than ever before.

One hundred and six thousand people will be diagnosed with colon cancer this year. More than 52 thousand will die from it. The best way to protect yourself — a colonoscopy.

Now doctors are using AI to help detect cancerous polyps.

“GI genius is an artificial intelligence technology that pulls from a pool of about 13 million pictures of polyps in real time and helps to identify high risk tissue in the colon during the procedure,” Nate Merriman, M.D., gastroenterologist at Intermountain Health, said.

GI genius works in tandem with the traditional colonoscopy. As the endoscopist examines the colon, the GI genius analyzes the images captured during the procedure to help identify suspicious spots.

“It puts a green box around the area as a whole that could have a polyp tissue in it. So, you can, in some ways, even see around corners or folds where you see a corner of the polyp,” Merriman said.

A recent study found that the GI genus scan showed a 14 percent increase in the detection rate for each physician. In the Intermountain Health System, Merriman says it could save as many as 18 lives every year.

“That’s significant over just five centers. You imagine how many colonoscopy centers we have in the country, how many patients we could save with this,” Merriman said.

Doctors agree early detection is the most important thing they can do to prevent colon cancer and death. It costs up to $250,000 to treat a patient once diagnosed with later-stage colon cancer.