Medical School OWL

The Medical Assistance Systems working group focuses on the user-centred design, development and evaluation of intelligent medical assistance systems for diagnostics, therapy, rehabilitation, care support and decision support for doctors. The intelligent assistance systems used range from motivating and therapy-accompanying apps to decision support systems and interactive agents for therapy support through to embodied robots.

Research is focussed on various areas. One of these is dialogues between agents and humans, with research here focusing on adapting to the human interaction partner by monitoring and directing their attention through verbal cues. Another focus is the examination of social, comprehensible artificial intelligence. When interacting with AI, people without in-depth knowledge of AI need explanations that are tailored to their individual needs. With TRR 318 "Constructing Explainability", we are developing approaches to take the specific needs of users into account. The third focus is system development related to stakeholder empowerment. The WG develops systems that are not only resilient, but can also be used and configured by non-experts by enabling users to program the behaviour of robots or agents according to the requirements of their application. Relevant stakeholders are involved in the process.

News

• Paper zu SHIFT wurde angenommen bei der IEEE ICDL 2025
• Erfolgreicher Auftakt für neuen Fokusbereich AI*IM an der Universität Bielefeld
• New Podcast episode “Scaffolding” is out now

Good to know

Showroom Open Hour: from 2 July every Wednesday from 12-13h.

 

AG MAS open source software now available on Docker Hub!

Good news for developers and open source fans: Our software is now available on Docker Hub! 🚀

You can find our containers at https://hub.docker.com/u/unibimas and try them out directly before integrating them into your projects. This makes using and deploying our open source solutions even easier and more efficient.

In the future, more projects will be available via Docker - so take a look, test the images and give us your feedback! 💡🔧