From April 28 to 30, 2026, researchers, clinicians, and data scientists gathered at Schloss Rauischholzhausen for the DZL DataLung Symposium. Set in the calm and historic surroundings of the castle, the meeting offered three days of focused scientific exchange, collaboration, and community building. For many, it was both a happy reunion of the established DataLung cohort and a warm welcome for new members joining the network.
The symposium opened with remarks by Dr. Christian Kalberlah, Managing Director of the German Center for Lung Research (DZL), followed by a welcome address from DataLung Chair Dr. Malte Lücken and Co-Chair Dr. Svenja Gaedcke. Together, they emphasized the growing importance of interdisciplinary work at the intersection of medicine, artificial intelligence, and data science - and the need to translate innovation into practice.
The scientific program reflected the breadth and depth of current AI-driven biomedical research. A central highlight was the keynote lecture “Large Models in Research AI Consulting: Promise, Pitfalls, and Practice” by Dr. Marie Piraud. Her talk offered a balanced view on the potential of large AI models, while also addressing practical and methodological challenges in real-world research settings. The panel discussion “LLMs for Research: Risks and Rewards” provided space for critical reflection. Panelists shared diverse perspectives on the transformative potential of large language models, while also addressing their limitations, risks, and ethical implications in scientific work.
Clinical applications of AI were another strong focus. Dr. Urs Eisenmann presented his work on AI-assisted MRI analysis of airway abnormalities and mucus plugging in cystic fibrosis. Prof. Dr. Andreas Mayr followed with “When Modern Prediction Models Meet Clinical Reality: Simplicity Matters,” arguing convincingly that usability and clarity are often more valuable than technical complexity in clinical prediction models.
A different impulse came from Dr. Jan Fuge’s and Dr. Raphael Majeed’s lecture “Vibe Coding.” The talk sparked lively discussion and led seamlessly into interactive group activities that encouraged informal exchange, collaboration, and new connections across disciplines and career stages. In addition to the formal sessions, poster presentations and networking activities created valuable opportunities for in-depth conversations and the start of new collaborations.
A particularly engaging ice-breaker and networking activity organized by Dr. Stephanie Hirn fostered lively exchanges among participants and helped strengthen connections across the community. The unique atmosphere of Schloss Rauischholzhausen provided an ideal setting to strengthen existing connections and build new ones within the DataLung community.
The program also featured Lightning Talks by new members, offering concise insights into emerging ideas and ongoing work across the network. A special community highlight showcased a success story originating from last year’s event: Project JUNO – From the Dancefloor into the Lab, presented by Julia Rückoldt and Noah Heldt.
The DZL extends its sincere thanks to the DZL DataLung Steering Board (Dr. Doreen Franke, Dr. Svenja Gaedcke, Dr. Jan Fuge, Dr. Raphael Majeed, Prof. Dr. Inke König, Dr. Malte Lücken, and Prof. Dr. Rory Morty) for their excellent organization and commitment, which made the symposium both productive and welcoming. Once again, the DataLung Symposium showed how much can be gained when diverse expertise comes together. It highlighted not only the progress of data-driven lung research, but also the strong, collaborative spirit that continues to define the DataLung School and its wider network.
With the DataLung School, the DZL specifically supports early-career researchers at the interface of pulmonary research, data science, and artificial intelligence. The program combines structured training in areas such as statistics, programming, and machine learning with a strong interdisciplinary network across the DZL.
