The best of artificial intelligence in 2024 (Gastrointestinal Endoscopy)

Artificial intelligence (AI) is transforming gastroenterology (GI) endoscopy, particularly in colonoscopy, polyp detection, and lesion classification. Recent studies demonstrate that AI enhances novice endoscopists’ accuracy, with randomized trials in China and Japan showing reduced adenoma miss rates when AI-assisted colonoscopy is used. Systematic reviews confirm AI’s effectiveness in detecting sessile serrated lesions, though AI’s impact on advanced adenomas is less clear. AI also supports training by improving learning curves and allowing less experienced endoscopists to match expert performance. Advances in AI-driven systems for polyp classification, quality control documentation, and Barrett’s esophagus detection are further broadening AI’s clinical utility. Additionally, AI in endoscopic ultrasound (EUS) aids in recognizing normal anatomy, potentially reducing learning time for challenging procedures. While regulatory limitations persist, platform-based systems and image repositories like GastroNet are expected to streamline AI integration, democratize expertise, and enhance efficiency in endoscopy. This review reflects AI’s rapid advancement and its anticipated role in standardizing and optimizing GI endoscopy.