University hospitals will use AI to screen for colorectal cancers in underserved communities

University hospitals will use AI to screen for colorectal cancers in underserved communities

Since March, doctors at the main campus of University Hospitals (UH) near University Circle in Cleveland have been using artificial intelligence to detect polyps during colonoscopies.

This fall, the hospital system will receive four additional units through a grant from Medtronic’s Health Equity Assistance Program to provide colorectal screening in medically underserved communities.

During the procedure, a green box appears on the monitor and highlights areas the AI ​​software thinks are polyps, said Dr. Gerard Isenberg, director of medical quality at the UH Digestive Health Institute.

He said recent studies show that doctors miss about 25% of precancerous polyps during colonoscopies.

“A 30-minute colonoscopy is about 54,000 images,” Isenberg said. “You can imagine a person looking at 54,000 images distracting you. Your eye sort of glosses over things.

He said the AI ​​program is like having a second set of eyes that almost never misses anything.

“This technology is one of the few things we have that can make a difference in reducing colon cancer,” Isenberg said.

The modules, which were cleared by the Food and Drug Administration last year, resulted in an almost 50% reduction in missed colorectal polys compared to a standard colonoscopy in the first US trial, according to UH.

If left undetected and untreated, polyps can lead to colorectal cancer. The new AI modules have a sensitivity rate per lesion of more than 99%, according to the hospital.

When detected early, some types of colorectal cancer have a five-year survival rate of over 90%, but the cancer remains the third most common and second deadliest in the country, according to UH.

Colorectal cancer is especially deadly in northeast Ohio where the death rate exceeds the national average. Diagnoses of the disease among Northeast Ohioans under age 50 are on the rise, according to UH.

“This statistic is worrying and shows why we need to start screening at age 45 in people without risk factors, as several national organizations now recommend, and even earlier in people with risk factors,” Isenberg said.

People can contact university hospitals to make an appointment for a colonoscopy without a doctor’s referral. There are options available for people without medical insurance, according to UH.

Isenberg said the hospital system plans to educate communities about the new technology by hosting screening events. Doctors will go to religious groups and community centers where there are disparities in income and access to care, which impacts who gets the colonoscopies.

“It’s not just about having this technology.” he said. “He’s trying to raise awareness about the dangers of colon cancer and why there’s such a disparity between whites and blacks.”

Black Americans have a lower survival rate five years after diagnosis, he said.

Anyone over 45 should get tested, according to Isenberg. People with inflammatory bowel disease or who have a family member diagnosed with colon cancer should be screened even earlier.

He said people shouldn’t be afraid of discomfort during a colonoscopy.

“With the ability to sedate these days, the new endoscopes we have, and the fact that people are better trained than maybe even 20 years ago, the procedure is actually very safe and easy,” Isenberg said.

The new units will be placed in every general endoscopy room on UH’s main campus, according to the hospital. Another will be transferred to UH’s Ahuja Medical Center in Beachwood.

“By finding and removing a polyp, we can prevent cancer from forming,” Isenberg said. “This AI technology helps us reduce variability in finding these precancerous polyps and has an impact on improving the health of our patients.”

Copyright 2022 WKSU. To learn more, visit WKSU.

#University #hospitals #screen #colorectal #cancers #underserved #communities

Leave a Comment

Your email address will not be published. Required fields are marked *