Aim

All routine gastric biopsies are screened for the presence of Helicobacter bacteria. Helicobacter infections, mainly by H. pylori, may cause gastric or duodenal ulcers and are associated with the development of gastric carcinoma and gastric lymphoma. Detection of the bacteria by conventional light microscopy in gastric tissue remains the gold standard for the diagnosis but can be challenging and time-consuming. In addition, our own preliminary results demonstrate that detection of Helicobacter bacteria on whole slide images is more challenging compared to conventional microscopy. Aim of the project is to generate an image analysis algorithm using deep learning to automatically detect the presence of Helicobacter bacteria in whole slide images in gastric biopsies.