Endoscopy has revolutionized gastrointestinal diagnostics, yet endoscope design has seen little change in 40 years, limiting effectiveness. This project introduces advanced flexible robotic endoscopic tools with digital control, improved maneuverability, and robotic training systems. These innovations promise precise submillimeter dissection and enhanced skill transfer through simulators. The project is a collaboration between University Hospital Dresden, TU Dresden, Leibnitz Institute, Contronix, Sitec Industrietechnologie, and Wolfram Designers, funded by EKFZ, Sächsischen Aufbaubank, and Volkswagenstiftung.

The enDDo.AI project leverages state-of-the-art AI methods, including foundation models, to analyze videos from various endoscopic domains. Our primary goal is to develop clinically useful tools that reduce physicians' workloads, prevent complications, and ultimately improve patient care during endoscopic procedures. Ongoing projects focus on the analysis of video capsule endoscopy and endoscopic retrograde cholangiopancreatography (ERCP), aiming to enhance diagnostic accuracy, assist in decision-making, and improve procedural efficiency in these specialized endoscopic domains.

The interdisciplinary team of researchers from the Department of Gastroenterology at Dresden University Hospital and the Chair of Microsystems at Dresden University of Technology collaborate on the ESF Plus project MultiMOD. The group addresses the extreme functional integration and miniaturization of optical fiber endoscopes with artificial intelligence by integration of kinaesthetic and tactile sensor systems, the actuation of the endoscope tip and the realization of path finding, automatic guidance and chemical senses.

Endoscopy training is a time-consuming and non-standardized process. Unlike laparoscopic surgery, there are no autonomous training courses available for prospective endoscopists to train independently. Therefore, we aim to develop independent endoscopy training courses with the help of software, allowing for independent practice.

Endoscopic procedures face complications due to navigation challenges. Optical methods have limitations, but radar-based imaging offers improvements. Integrating radar in endoscopes enhances navigation, reduces injury risks, and improves outcomes. This project, a collaboration between University Hospital Dresden, the University of Dresden's High-Frequency Technology Chair, and Vmed, aims to advance transmural imaging capabilities.

As part of BMBF Cluster4Future SEMECO, partners are developing an intelligent capsule endoscope with adaptive anchoring, biopsy capabilities, and a broadband chemical sensor for in vivo microbiome mapping, involving University Hospital Dresden, TU Dresden, Renesas Dresden, and Ovesco.

The aim of SedAi (also known as Sedation AI) is to use objective vital parameters to enable automated sedation monitoring with a reduction in complications during endoscopic procedures. Vital parameters are recorded during these procedures. This project involves several institutions across Germany, including Dresden, working together under the leadership of the University Hospital Halle to automate and improve sedation in endoscopy.

Our clinical endoscopic research focuses on advanced procedures like ERCP and EUS interventions. For instance, we've investigated bleeding complications from draining walled-off necroses with lumen-apposing metal stents. Additionally, we've studied bile duct stent migration rates and conducted material analyses to uncover factors influencing long-term success. Currently, we're exploring EUS-guided biopsies and percutaneous endoscopic gastrostomy techniques.