Artificial intelligence is rapidly transforming gastroenterology — from the endoscopy suite to the clinic. Experts highlight how AI models are moving beyond polyp detection to tackle some of the field’s toughest diagnostic challenges.
Drs. Yuvaraj Singh, Alessandro Colletta, and Neil Marya showcased AI systems that outperform traditional ERCP sampling in identifying malignant biliary strictures and can even distinguish autoimmune pancreatitis from pancreatic cancer with higher accuracy than experts. These breakthroughs, trained on millions of images, promise faster, more reliable diagnoses in complex pancreatic and biliary disease.
At a broader level, Dr. Dennis Shung discussed how foundation models like GastroNet-5M are redefining endoscopy — from “digital biopsies” to AI-generated reports and clinical decision support. Yet, he cautioned that overreliance could erode clinician skill and patient connection. Professional societies such as the AGA are now crafting “living guidelines” to ensure AI integration remains ethical, transparent, and patient-centered.