In a report by Amy Pfeiffer, a large randomized trial found that AI assistance during upper endoscopy did not significantly improve detection of gastric neoplasms after centralized pathology review.
The study, led by Zehua Dong, PhD, and published in Gastroenterology, analyzed 29,514 patients across 24 hospitals in China comparing standard EGD with AI-assisted EGD using the ENDOANGEL-GN system. Detection rates were 1.42% with AI vs 1.25% without AI, a difference that was not statistically significant after expert pathology review.
However, AI reduced blind spots during endoscopy and increased inspection time, and subgroup analysis suggested potential benefits for less experienced endoscopists or during fatigue periods.