Is AI-assisted colonoscopy ready for clinical practice?

The ‘GI Genius’ artificial intelligence (AI) endoscopy module improved the adenoma detection rate (ADR) and sessile serrated polyps (SSPs) detection rate in patients undergoing colonoscopy. According to the authors, the results indicate that the ‘GI Genius’ system could be routinely used in most patients undergoing this procedure to improve polyp detection rates and, consequently, to reduce post-colonoscopy colorectal cancer incidence.

“Polyp detection rates are correlated with post-colonoscopy colorectal cancer rates,” outlined Dr Alexander Seager (Newcastle University, UK) [1]. “Therefore, any increase in detection may result in a direct patient benefit.” The current randomised-controlled trial compared the effectiveness of ‘GI Genius’-assisted colonoscopy (GGC) with standard colonoscopy (SC) in routine practice. For this purpose, 2,032 participants who were scheduled for colonoscopy were randomised 1:1 to GGC or SC. The study was executed across 10 hospitals in England. The primary endpoint was mean adenomas per procedure (MAP).

The MAP rates were 1.21 for participants in the SC group and 1.56 for participants in the GGC group (IRR 1.30; 95% CI 1.15–1.47; P<0.001). The key secondary outcome also favoured the GGC arm, with ADRs of 0.48 and 0.57 for the SC and GGC arms, respectively (OR 1.47; 1.21–1.78; P<0.001). “We also saw an SSP detection rate increase of 3.3% with GGC,” said Dr Seager. “If we bear in mind that a 1% increase in proximal SSP detection rate can decrease post-colonoscopy colorectal cancer rate by 7%, this is potentially a big finding.”

As expected, adjusted mean polyp sizes were smaller in the GGC arm, with 5.18 mm, compared with 5.78 mm in the SC arm (P=0.016). Interestingly, the study also showed that GGC was associated with an increased detection rate of polyps of a specific morphology, namely 0-IIa polyps (34% vs 27%; P=0.002). Finally, there were no significant differences with respect to withdrawal time or safety events.

In conclusion, GGC outperformed SC in terms of overall polyp detection rate, as well as ADR and SSP detection rate, without increasing adverse events. The increased detection rate for SSPs, in particular, may strongly reduce post-colonoscopy colorectal cancer rates.

1. 
   
   1. Seager A, et al. Colo-detect: a randomised controlled trial of polyp detection comparing colonoscopy assisted by the GI genius artificial intelligence endoscopy module with standard colonoscopy. LB18, UEG Week 2023, 14–17 October, Copenhagen, Denmark.

Copyright ©2023 Medicom Medical Publishers



Posted on 7 December 20237 December 2023