AI-assisted GI Genius increases ADR, improves sessile serrated polyp detection by 3%
Click Here to Manage Email Alerts
Key takeaways:
- Use of GI Genius increased mean adenomas per colonoscopy (1.56 vs. 1.21) as well as adenoma detection rate (56.6% vs. 48.4%).
- The module also had a higher sessile serrated polyp detection rate (11.6% vs. 8.3%).
An artificial intelligence endoscopy module increased the detection of adenomas, including sessile serrated polyps, during colonoscopy and is recommended for implementation into routine practice, a presenter said at UEG Week.
“Increasing [polyp] detection matters,” Alexander Seager, of South Tyneside and Sunderland NHS Foundation Trust, said. “All of our standard key performance indicators for detection are inversely correlated with [post-colonoscopy colorectal cancer (PCCRC)] rates and there’s no known ceiling effect for increasing that detection. Theoretically, if we can increase detection, even to a small degree, we may then get a benefit in reducing PCCRC.”
In the multicenter, randomized controlled COLO-DETECT trial, Seager and colleagues used the GI Genius intelligent endoscopy module (Medtronic) to investigate the clinical effectiveness of AI-assisted colonoscopy compared with standard colonoscopy for screening (60.6% for bowel cancer) or gastrointestinal symptoms/surveillance.
Of 2,032 participants recruited (mean age, 62 years; 56% men), 1,015 underwent GI Genius-assisted colonoscopy (GGC) and 1,017 underwent standard colonoscopy. The primary outcome was mean adenomas detected per procedure.
Results showed a significant gain in mean adenomas per colonoscopy with GGC vs. standard care (1.56 vs. 1.21; incident rate ratio = 1.3; 95% CI, 1.15-1.47), as well as a higher adenoma detection rate (56.6% vs. 48.4%; OR = 1.47; 95% CI, 1.21-1.78). GGC also identified smaller polyps (mean, 5.18 mm vs. 5.78 mm; 95% CI, –1.08 to –0.11), more 0-IIa polyps (34% vs. 27%), and, notably, had a higher sessile serrated polyp (SSP) detection rate (11.6% vs. 8.3%; OR = 1.49; 95% CI 1.09-2.03).
“If we bear in mind that a 1% increase in proximal sessile polyp detection rate can decrease your PCCRC rate by 7%, this is potentially a really big finding,” Seager said.
Further, overall polyp count was higher with GGC (2.23 vs. 1.65; IRR = 1.35; 95% CI, 1.21-1.5), as was polyp detection rate (OR = 1.56; 95% CI, 1.29-1.9). Procedural time was similar between the two groups (23.97 min. vs. 22.5 min), although in the absence of polyps, procedures took slightly longer overall, Seager noted.
“The COLO-DETECT trial showed significant increases and clinically significant as well as statistically significant increases in adenomas and SSPs, those are smaller and flatter, with a small increase in procedure time and no increased adverse events between the two arms,” Seager concluded. “Based on our results, we can recommend the GI Genius could be adopted into routine colonoscopy practice ... to improve polyp detection and, therefore, to potentially decrease post-colonoscopy colorectal cancer rates.”
Collapse
Seager A, et al. A. randomized controlled trial of polyp detection comparing colonoscopy assisted by the GI Genius artificial intelligence endoscopy module with standard colonoscopy. Presented at: UEG Week; Oct. 14-17, 2023; Copenhagen, Denmark (hybrid meeting).
Read more about
Click Here to Manage Email Alerts