AI supports diagnostics during gastroscopies

Researchers from Julius-Maximilians-Universität Würzburg and Charles University in Prague are working together to develop an AI that can detect diseases during gastroscopy and better assess the need for tissue samples.

The Bavarian Ministry of Science is funding the Würzburg part of the project with EUR 224,000 as part of the "Joint Czech-Bavarian Research Projects 2024-2026" program. GI-Insight will run for two and a half years, with the aim of developing a product prototype by the end of 2026 that can identify diseases in real time during a gastroscopy.

"In order for an AI to reliably recognise and classify abnormalities based on images from the stomach, it has to be trained with a lot of data," says Adrian Krenzer. Annotating the training data is a particular challenge here. Pavel Pecina uses natural language processing (NLP) methods to link extensive written medical findings with gastroscopic images and collect them in a database. This annotated data is then used by Adrian Krenzer's team to train the AI.

Once the AI prototype has been developed, it will be tested at Würzburg University Hospital. Alexander Hann, Professor of Digital Transformation at the Chair of Gastroenterology at the University Hospital of Würzburg, will be responsible for the medical validation of the AI results. Hann is already cooperating with the JMU Chair of Artificial Intelligence and Knowledge Systems led by Professor Frank Puppe.

The GI-Insight project shows how artificial intelligence can support medical diagnostics by increasing the accuracy of gastroscopies and reducing the workload of doctors. International collaboration and funding will significantly advance research in this field.