Artificial Intelligence Makes Colonoscopies Smarter

December 21, 2018 By IdeaConnection

AI Makes Colonoscopies SmarterArtificial intelligence is spreading into all areas of our lives.  It is also reaching parts of our bodies helping to save our lives.

Doctors at the University of California Irvine are using AI to improve colonoscopies.  The system operates similarly to facial recognition.

A colonoscopy is the most effective way to diagnose bowel cancer.  A flexible tube with a camera and light at its tip is inserted into the body to examine the bowel for any abnormalities such as polyps. These are small growths on the inner lining of the rectum or the large intestine (colon).   Polyps don’t usually turn into cancer but if some types aren’t removed there is a chance they could become cancerous.

The artificial intelligence system developed by doctors at UCI has been designed to spot polyps that might be tricky for colonoscopists to discover first time around.

According to Dr. William Karnes who helped to develop the system it has the ability to identify up to 20 percent more abnormalities than the average, traditional test.  It uses complex algorithms and analyzes more than 98 images per second as it searches for polyps and candidate tissue that could become one.

This new artificial intelligence approach presents its results to doctors in real-time during the colonoscopy procedure.

Great Results

Ultimately the AI system can improve the rate of detection.

In a March press release, Karnes spoke about the promising results from early testing of the system:

“We had three experts at UCI with very high ADRs (Adenoma detection rate) review the videos.

“We had them mark every polyp they saw. On first review, they found about 20 percent more polyps than the original colonoscopist had removed. Then they viewed it with an AI overlay; AI found all the polyps they had discovered, plus about 20 percent more polyps than even they did.”

Currently, doctors at UC Irvine are offering the AI colonoscopy one day a week.



Share on      
Next Post »

Add your Comment

[LOGIN FIRST] if you're already a member.

fields are required.

Note: Your name will appear at the bottom of your comment.