Post written by Jason B. Samarasena, MD, from the H. H. Chao Comprehensive Digestive Disease Center, Division of Gastroenterology & Hepatology, Department of Medicine, University of California, Irvine, Orange, California, USA.
Esophageal adenocarcinoma is actually rising in this country and not going down! The visual detection of early esophageal neoplasia in BE with white-light and virtual chromoendoscopy still remains challenging. Endoscopists may be missing neoplastic lesions during BE surveillance exams given they are often subtle in nature. Furthermore, many endoscopists are not adequately trained to detect these lesions. We felt that artificial intelligence could augment an endoscopist’s skill set in detecting neoplasia in BE.
Our AI algorithm correctly detected early neoplasia with a sensitivity of 96.4%, specificity of 94.2%, and accuracy of 95.4% on still images. In addition, the object detection algorithm was able to draw a localization box around the areas of neoplasia with high precision and at a speed that allows for real-time implementation.
This study sets the stage for development of a real time video-based AI algorithm that can assist endoscopists detect neoplasia in Barrett’s Esophagus earlier so that preventative therapy can be carried out and cancer can be averted.
Figure 2. Illustration of the deep learning system. CNN, Convolutional neural network; NMS, non-maximum suppression.
Read the full article online.
The information presented in Endoscopedia reflects the opinions of the authors and does not represent the position of the American Society for Gastrointestinal Endoscopy (ASGE). ASGE expressly disclaims any warranties or guarantees, expressed or implied, and is not liable for damages of any kind in connection with the material, information, or procedures set forth.