Real-time AI-assisted Medical Diagnosis on Gastrointestinal Endoscopic Videos
, Chulalongkorn University
Endoscopy is a procedure used in medicine to look inside the body. The procedure uses an endoscope to examine the interior of a hollow organ or cavity of the body. For gastrointestinal (GI), there are two common operations: colonoscopy and gastroscopy. In this project, we aim to implement a deep learning model to help doctors diagnose in real time during the endoscopy operation. First, the object detection model aims to detect polyps on colonoscopy video for early detection of colorectal cancer. Second, the semantic segmentation model aims to segment the gastric intestinal metaplasia areas for an early detection of gastric cancer.