Institute of Formal and Applied Linguistics
Charles University, Czech Republic
Faculty of Mathematics and Physics
Deep learning enables enormous progress in many tasks in the field of image processing. Artificial intelligence (AI) continually delivers new top results in the areas of detection and classification, sometimes matching or even exceeding human performance. These achievements impact many areas, including medical applications. A special area of medical applications is gastroenterology. Machine learning algorithms are already being used in gastroenterology to support gastroenterologists during interventions. Gastroscopy is a commonly performed medical procedure that is critical to the diagnosis and treatment of many gastroenterological diseases. Despite its frequency, gastroscopy presents challenges ranging from the identification of a variety of different findings to the need for extensive medical knowledge and experience. Particularly for medical beginners, the complexity of the required diagnoses can lead to errors and
incorrect treatment approaches. Our project aims to improve the precision and safety of gastroscopy by empowering medical professionals with advanced AI technologies. The project combines deep learning algorithms with extensive medical expertise from previous gastroscopy examinations (medical reports) to perform image analysis during examinations. This enables the collection and analysis of relevant diagnostic features that are essential for accurate assessment and treatment.