KILABS: Interaktiver Antibiotic-Stewardship-Leitfaden in Form einer Kl-basierten dialogfähigen Anwendung für die Behandlung von Infektionskrankheiten

Dr. Tobias Depke , Dr. Friedhelm Bach , Dr.-Ing. Christoph-Alexander Holst
01.07.2026 bis 31.12.2027

Intelligent Assistance for Clinical Practice

In the KILABS project, inIT is collaborating with the Applied Health Science Group at TH OWL and the Evangelisches Klinikum Bethel to develop an AI-powered assistance system for everyday clinical practice.  The system is designed to help clinicians find relevant information on the diagnosis and antibiotic treatment of patients with infectious diseases quickly and reliably.

To achieve this, specialised language models are combined with evidence-based infectious disease guidelines and local clinical recommendations. The interactive Antibiotic Stewardship guide will be accessible through both text and speech and will support healthcare professionals directly at the point of care. It is not intended to replace medical decision-making, but to complement it by providing straightforward access to professionally reviewed information.

Developed in Close Collaboration with Clinical Practice

Alongside the technical development, practical testing in everyday clinical settings is a central part of the project. In collaboration with the EvKB, the requirements and experiences of future users are incorporated into the development process from an early stage. This ensures that the assistance system is aligned with actual hospital workflows and the information needs of healthcare professionals.

Project Objectives

The aim of KILABS is to develop a demonstrator that provides a dialogue-based access to relevant guidelines and clinical recommendations for important infectious diseases. An intuitive text-based interface is intended to help users find the information they need more quickly, support clinical processes and reduce the workload of healthcare professionals.

KILABS within the ISyCARE Innovation Community

As part of the ISyCARE innovation community, KILABS contributes to improving the care of patients with infectious diseases and to the responsible digital transformation of healthcare. The project combines the inIT’s expertise in artificial intelligence with medical knowledge and practical experience from everyday clinical practice.

The project focuses on supporting healthcare professionals, improving clinical processes and strengthening collaboration between research and practice. At the same time, KILABS is establishing a foundation for knowledge transfer within the regional healthcare network and for the future use of specialised language models in further medical applications.

This project is promoted by:
Federal Ministry of Research, Technology and Space (BMFTR) - DATIpilot
Sponsors: Projektträger Jülich
Funding Lines: ISyCARE Communityprojekt
Stakeholders / Contacts: Dr.-Ing. Christoph-Alexander Holst
Promoted by
Project sponsor