Artificial Intelligence in Gastrointestinal Endoscopy

Artificial intelligence (AI) systems based on machine learning have evolved in the last few years with increasing applicability in gastrointestinal endoscopy. Gastrointestinal endoscopy has proved to be a perfect context to develop such artificial intelligence (AI) systems that can aid endoscopists in many tasks of their daily activities. Artificial intelligence is expected to significantly influence the practice of Gastrointestinal Endoscopy in the near future. The clinical applications of AI relevant to GI endoscopy include lesion detection (computer-aided detection، CADe) and lesion characterization (computer-aided characterization، CADx). Deep learning techniques such as convolutional neural networks have been used in several areas of GI endoscopy، including colorectal polyp detection and classification، detection and depth assessment of early gastric cancer، dysplasia in Barrett’s esophagus، and detection of various abnormalities in wireless capsule endoscopy images. AI can help endoscopic physicians improve the diagnosis rate of various lesions، reduce the rate of missed diagnosis، improve the quality of endoscopy، assess the severity of the disease، and improve the efficiency of endoscopy.

Further research including randomized، multicenter trials are needed to further evaluate the use of artificial intelligence for real-time endoscopy. The aims of this review are to discuss the available AI-related findings and clinical applications in GI endoscopy، as well as to outline the current limitations and future directions in this field.