Scientists at Radboud teaching hospital in Nijmegen are using AI to detect polyps in the colon, and say the smart camera is leading to 40% more being found than with the naked eye alone. 

Polyps are small growths that can be a precursor of cancer. Up to now they have been identified and removed by colonoscopy and more than 100,000 people undergo such research every year. 

Previous studies have shown that the more polyps a doctor finds and removes, the lower the likelihood of developing colorectal cancer later on.

“A doctor can miss a polyp if, for example, it is only on the screen for a short time or if they are focused on another area,” researcher Michiel Maas said. “But a computer looks at every pixel and is alert every millisecond. And that has tremendous added value.”

The Radboud research, published in The Lancet Digital Health, was carried out across 10 hospitals in different countries and 916 patients took part. They had all either had polyps earlier or had been earmarked for colon cancer screening. 

Some were given regular checks and the rest used the AI method. The computer placed a purple square in the image if an area looked suspicious, drawing the doctor’s attention to it for closer examination.

Although the computer helped detect and remove more polyps, the colonoscopy with digital assistance did not take more time. Nor did the new procedure result in a higher percentage of false positives.

Next the researchers will investigate whether the computer can also determine the danger level of a polyp based on its visual characteristics and size.

“If we can determine that during the examination, we can leave harmless polyps in place,’ said gastroenterologist Peter Siersema. “They won’t need to be analysed by the pathology lab and ultimately, this will reduce the risk of complications and costs.”