Labelling Stomach Anatomical Locations In Upper Gastrointestinal Endoscopic Images Using a CNN

In this paper, we aim to develop a diagnostic assistant system for labelling the stomach anatomical locations in upper GastroIntestinal Endoscopy (UGIE) examination. To address this task, we construct an appropriate manner which utilizes both ability of a convolutional neural network (CNN) and supportive interactions between machine and doctors. We solve the problem by a two-phase scheme. The first is a coarse-phase to classify seven major anatomical locations including cardiac orifice, gastric body, fundic, antrum, pyloric ring, lesser curvature and greater curvature using advances of a CNN. The constructed CNN network is compact with high performance and appropriate integration into a Graphic User Interface (GUI). In order to classify with 13 further detailed positions, a GUI is developed so that the endoscopists can conveniently specify the anatomical locations from results of the coarse-phase. In this the fine-phase, the doctors will prune the automatic results as well as specify the more detailed positions of each major location. In the experimental results, the developed application is shown as an efficient way in an UGIE diagnosis. It reduces a significant time from averagely 13:03 minutes in a manual procedure to 4:35 minutes by using the developed system when comparing with trainee endoscopists. The results of specifying anatomical locations satisfied accuracy requirements and showed promising research trend for future application as a computer-aided GIE diagnostic system.