New insights on missed colonic lesions during colonoscopy through artificial intelligence–assisted real-time detection

Post written by Wai K. Leung, MD, from the Department of Medicine, Queen Mary Hospital, University of Hong Kong, Hong Kong.


This study aimed to determine whether the use of artificial intelligence (AI)-assisted real-time detection could reduce colorectal adenoma miss rate during colonoscopy.

Previous studies showed that up to 26% of adenomas could be missed during colonoscopy and potentially lead to interval cancer development. There is a pressing need to bring down the adenoma miss rate of colonoscopy. With the availability of real-time AI-assistance, we determined whether AI could reduce the adenoma detection rate and the underlying mechanism leading to missed adenomas.

Our study was divided into 2 parts. Part I involved the re-examination of the video recordings of 65 tandem examinations of the proximal colon by our validated real-time deep learning AI model. We found that 79.1% of missed proximal adenomas found on tandem examinations could indeed be detected by the AI in the video of the first-pass examination. Part II included 52 prospective colonoscopy examinations in which the endoscopists were blinded to the real-time AI findings. Segmental unblinding of the AI findings were provided to the endoscopists, and those segments would be reexamined if any additional polyps were detected by the AI but not the endoscopists. In this prospective examination, AI could detect at least 1 missed adenoma in 26.9% of patients and increased the total number of adenomas detected by 23.6%. Multivariate analysis showed that the presence of multiple polyps and colonoscopy by less experienced endoscopists are the 2 factors associated with missed adenomas. Together, we found that real-time AI could reduce up to 80% of missed adenomas during colonoscopy examination.

Although AI could possibly reduce up to 80% of missed adenomas, there are still about 20% of adenomas that are not detected even with AI-assistance. Future studies should examine the adjunct role of combined approaches with AI, such as tandem examination and use of distal attachment device, in eliminating missed adenomas during colonoscopy.


Figure 3. Typical appearances of missed adenomas picked up by AI (upper row) and the corresponding photo for the removal of them during re-examination (lower row).

Read the full article online.

The information presented in Endoscopedia reflects the opinions of the authors and does not represent the position of the American Society for Gastrointestinal Endoscopy (ASGE). ASGE expressly disclaims any warranties or guarantees, expressed or implied, and is not liable for damages of any kind in connection with the material, information, or procedures set forth.

Leave a Comment

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s