Computer-Assisted Detection Improves Identification of Proximal Colonic Lesions

Computer-assisted detection can improve proximal colonic polyp and adenoma detection.

Use of computer-assisted detection (CADe) improved detection rates of polyps and adenomas in the proximal colon but did not significantly reduce proximal colonic lesion miss rates, according to study findings published in Gastrointestinal Endoscopy.

Researchers conducted a prospective, multicenter, randomized controlled trial ( Identifier: NCT04294355). A total of 216 adult patients (mean age [SD], 61.6[10]; men, 48.6%) were randomly assigned 1:1 to undergo tandem examination of the proximal colon with CADe or conventional white light colonoscopy. Of the study participants, 54.7% underwent colonoscopy for routine screening, while the remaining presented with symptoms indicating the need for colonoscopy.

Following each of these procedures, clinicians performed a tandem examination of the proximal colon with white light in both groups. This enabled clinicians to calculate the adenoma/polyp miss rate based on lesions identified in the second examination that were overlooked during the first examination.

The overall adenoma detection rate totaled 48.9%, with 7.1% advanced adenomas and 1% cancer detection. The researchers did not observe any significant difference between the 2 groups in miss rates for proximal colon adenomas (CADe vs conventional: 20% vs 14%; P =.07) or polyps (CADe vs conventional: 26.7% vs 19.6%; P =.06).

This multi-center trial from Asia confirmed that CADe can further enhance proximal adenoma and polyp detection, but may not be able to reduce missed proximal colonic lesions.

In contrast, CADe detected proximal colon polyps and adenomas at significantly higher rates than the conventional white light colonoscopy (58% vs 46.7%; P =.03 and 44.7% vs 34.6%; P =.04, respectively). CADe also detected a higher average number of proximal colon polyps (1.20 vs 0.86; P =.03) and adenomas (0.91 vs 0.68; P =.03) compared with the conventional group.

Factors influencing higher adenoma detection rates included use of CADe, highly skilled endoscopists with higher baseline adenoma detection rates, superior bowel preparation, patients undergoing screening endoscopies, and a withdrawal time after 5 or more 5 minutes in the proximal colon.

CADe benefited patients undergoing screening endoscopies more than those presenting with symptoms necessitating the endoscopy. CADe also detected more adenomas and polyps in procedures that took less than 5 minutes prior to withdrawal from the proximal colon.

“This multi-center trial from Asia confirmed that CADe can further enhance proximal adenoma and polyp detection, but may not be able to reduce missed proximal colonic lesions,” the study authors noted.

Study limitations include potential overestimation of CADe effects in reducing missed lesions while estimating sample size, lack of blinding using the same endoscopist for the second tandem examination leading to potential bias, and lack of separation of screened vs symptomatic patients.

Disclosures: Some study authors declared affiliations with biotech, pharmaceutical, and/or device companies. Please see the original reference for a full list of authors’ disclosures.


Lui TK, Hang DV, Tsao SK, et al. Computer-assisted detection versus conventional colonoscopy for proximal colonic lesions: A multi-center, randomized, tandem colonoscopy study. Gastrointest Endosc. Published online October 5, 2022:S0016-5107(22)02018-1. doi:10.1016/j.gie.2022.09.020