7th Global Gastroenterology and Artificial Intelligence Summit (Live)

12 - Akshintala_Updated

Pdf Summary

This document summarizes recent advances and future directions of artificial intelligence (AI) applications in gastroenterology (GI), emphasizing computer vision in endoscopy, risk prediction, decision support, ambient AI, and workflow automation. AI in GI has seen exponential growth in research, with key breakthroughs in colonoscopy for polyp detection and classification attaining accuracies up to 96.4%. AI tools have also improved diagnosis in Barrett’s esophagus, early gastric cancer, and malignant biliary strictures, often employing advanced imaging modalities such as narrow-band imaging (NBI) and cholangioscopy.<br /><br />Capsule endoscopy reading times have been significantly reduced using AI, although real-world evaluations raise concerns about sensitivity and specificity limitations in certain contexts. Endoscopic ultrasound (EUS) applications include structure and mass detection, classification of pancreatic cancers versus autoimmune pancreatitis, and texture analysis linked to pathology and progression risk prediction. Federated learning is highlighted as a promising approach to multi-institutional model training while preserving patient privacy.<br /><br />Future endoscopy envisions integration of robotics and AI to enable semi-autonomous “self-driving endoscopes,” along with autonomous robotic surgeries. Clinical decision-making is enhanced by predictive models for conditions such as choledocholithiasis and pancreatic cyst management (e.g., COMPCyst), improving accuracy beyond current standards. AI also aids in post-ERCP pancreatitis prophylaxis by identifying risk factors and guiding interventions.<br /><br />Ambient AI reduces non-endoscopy administrative tasks, saves time, and supports documentation through tools like EndoScribe, which uses speech recognition and language models for note generation and quality metric tracking. AI-driven scheduling optimizes endoscopy room efficiency by predicting cancellations and reducing wait times.<br /><br />The American Society for Gastrointestinal Endoscopy (ASGE) offers innovation and commercialization guidance and workshops to translate AI research into scalable, market-ready solutions. The take-home message stresses AI’s growing role in personalized diagnosis, treatment enhancement, efficiency gains, and the importance of moving beyond publications toward real-world applications and advocacy.

Keywords

artificial intelligence

gastroenterology

computer vision

endoscopy

colonoscopy

polyp detection

capsule endoscopy

endoscopic ultrasound

federated learning

robotic surgery