The team of the new Department for Medical AI and Translation (MedKIT) at UKL: Prof. Toralf Kirsten (top left), Martin Federbusch (top right), Maria Schmidt (bottom right), Dr. Daniel Steinbach (bottom center), Alexander Twrdik (left).

The University of Leipzig Medical Center (UKL) has set up a new department for Medical AI and Translation (MedKIT). Its team, which is part of the Medical Informatics Center, is dedicated to the implementation of medical AI applications in clinical practice. The primary goal of MedKIT is to find and implement scientifically sound AI solutions for patient care. It is building on the success of the AMPEL project which has been implementing real-time support in everyday inpatient care at the UKL for years. The task of the new department is to develop and operate clinical AI support at the UKL.

The MedKIT team has come out of the AMPEL project, thus allowing the extension of the UKL’s successful open source initiative and AI platform AMPEL-CDSS. Over the coming months, the current five-person team will be focusing on recording and pooling current AI initiatives at the UKL and on establishing itself as a central point of contact for the use of AI in patient care. “Our task is to find, test and implement the right scientifically sound AI solutions for the challenges our colleagues in the hospitals are facing,” says Prof. Toralf Kirsten, Head of the Medical Informatics Center. “In my opinion, the K in MedKIT clearly stands for collaboration. We need to move away from insular thinking and create an AI world where everyone can participate.”

To achieve this, the UKL is setting up the MedKIT department as a central hub for pooling all AI activities in patient care, while working closely with the Faculty of Medicine, where the relevant research activities are concentrated. “One of the goals we have set ourselves in our current digitalization strategy is to increase the use of AI in diagnostics and clinical decision support in order to further improve patient care,” says Dr. Robert Jacob, Chief Commercial Officer of the UKL.

AMPEL-CDSS is an excellent example of how this kind of collabo-ration, which is ‘bench-to-bedside’ in action, can succeed. “Our experience with the AMPEL project has shown that real innovation only happens by collaborating,” Martin Federbusch says. The specialist in laboratory medicine, internal medicine, and clinical decision support was project manager of the AMPEL project and is now head of MedKIT. In this role he is responsible for transitioning the former AMPEL project into routine patient care. “This is a translational step where most scientific projects fail,” he explains.

Decision support, which began in laboratory medicine, has made its way into almost all departments at the UKL. “Today, we can’t imagine life without AMPEL,” Federbusch says. He explains that the real benefits of AI only become apparent in a clinical setting. “Our aim now is to network with as many other departments and hospitals as possible, to share our experience, and to also collect innovations from others.” So the MedKIT team founded the CDSN (Clinical Decision Support Network), an informal network in which 50 interdisciplinary partners from across Germany come together to share knowledge, collaborate, and build joint infrastructures.

The AMPEL platform has also been established at Dresden University Hospital as part of a research project, with the aim of finding out whether and how it might be implemented at other hospitals. This project is being supported with funding from the Free State of Saxony amounting to around one million Euros.

Read more here: uniklinikum-leipzig.de