AI to aid colon exams
AI to aid colon exams

A NEWLY developed real-time artificial intelligence (AI) system will help doctors pick up colorectal polyps missed by human eyes.
This system was developed by the research team at the Chinese University of Hong Kong’s Department of Surgery (中文大學外科學系). By connecting colonoscopy images to the system, which has stored over 50,000 data samples, any suspected polyp areas will be circled on real-time images, enhancing their accuracy following a doctor’s diagnosis.
A colonoscopy is one of the most effective ways to perform colorectal cancer screenings, as it allows doctors to examine the entire colonic wall and remove any polyps if necessary.
But, according to the assistant professor of the department, Carmen Poon Chung-yan (潘頌欣), 10 to 20 percent of polyps are missed on average during the process. A polyp may develop into colorectal cancer.
The new system will not replace a doctor’s work in conducting colonoscopies. Rather, it aims to facilitate and assist the process by reducing the risk of failing to detect and classify colorectal polyps.
She said the installation cost of the system on each device is a five-digit amount. The research team expects to launch the system onto the market in one to two years.
The system was showcased at the International ICT Expo 2018 at the Hong Kong Convention and Exhibition Centre in mid-April.

(This article is published on Junior Standard on 18 April 2018)

CUHK develops a real-time diagnosis system for cancer screening

Benefits and risks of artificial intelligence