Professor Colin Rees
President of the British Society of Gastroenterology, Professor of Gastroenterology, Population Health Sciences Institute, Newcastle University Centre for Cancer, Faculty of Medical Sciences, Newcastle University, Paul O’Gorman Building
AI is reshaping how we detect and treat disease. In bowel cancer screening, it plays a vital role in spotting hard-to-see polyps, improving early diagnosis and outcomes.
Artificial intelligence in healthcare
Artificial Intelligence (AI) is revolutionising the way healthcare is delivered. It is being used in a variety of ways, including to look at large datasets, report scans, aid planning of treatment and improve existing technologies. AI also has a major role to play in how we detect cancer and abnormalities in the body that may turn into cancer.
Bowel cancer in the UK
Bowel cancer affects 42,000 people in the UK each year. Most bowel cancers begin as pre-cancerous lesions called polyps. Polyps are small, wart-like lesions on the wall of the bowel that, over time, can turn into cancer. This process of changing into a polyp, then into cancer may take as long as 10–15 years. If we can diagnose these polyps, we can remove them and stop them from turning into cancer.
Several research studies have
now demonstrated that AI detects
more polyps in the bowel.
How is bowel cancer detected?
Polyps are found through a camera test looking into the bowel, called a colonoscopy. A colonoscopy is used to detect cancer in the bowel and to pick up polyps. These polyps can sometimes be hard to spot, and they can be missed. AI technology has been developed to provide a second pair of eyes, to look thoroughly at the bowel and find polyps that may have been missed.
AI can help detect more polyps
Several research studies have now demonstrated that AI detects more polyps in the bowel. A large UK study conducted across 10 sites and published in the Lancet group of journals1 demonstrated that around 30% more polyps were found using AI. Other trials have shown similar results. This trial showed that AI found smaller, flatter polyps — the sort most likely to be missed during a colonoscopy.
The trial found more of a particular type of polyp, called sessile serrated lesions, which are harder to spot and more likely to progress rapidly to cancer. Using this AI technology more widely in the NHS will allow more polyps to be found and removed, with the potential for preventing cases of bowel cancer.
[1] Alexander Seager, Linda Sharp, Laura J. Neilson, Andrew Brand, James S. Hampton, Tom J.W. Lee, Rachel Evans, Luke Vale, John Whelpton, Nathania Bestwick, Colin J. Rees on behalf of the COLO-DETECT trial team. COLO-DETECT: A randomised controlled trial of polyp detection comparing colonoscopy assisted by the GI Genius™ artificial intelligence endoscopy module with standard colonoscopy in routine colonoscopy practice. The Lancet Gastroenterology & Hepatology, Volume 9, Issue 10, 911 – 92
