Universtitaet Bern
Founded in 1834, the University of Bern is one of Switzerland's leading comprehensive research universities, bringing together some 25,000 students and staff across eight faculties, more than 180 institutes and research centres, and a vibrant ecosystem that drives scientific excellence, innovation and societal impact.
Within the University, the ARTORG Center for Biomedical Engineering Research was established in 2008 to bridge engineering and medicine, and today comprises multidisciplinary research units working closely with clinical partners at the Inselspital, Bern University Hospital.
UBERN's contribution to AIRIS is led by the Artificial Intelligence in Health and Nutrition (AIHN) Laboratory at the ARTORG Center, which develops AI-based approaches that integrate multimodal and longitudinal health data with biomedical knowledge to enable personalised monitoring, diagnosis, disease progression assessment and management. AIHN has particular expertise in developing trustworthy and mechanism-informed AI for complex chronic diseases, with a focus on metabolic and lung diseases.
The Department of Health Psychology and Behavioral Medicine of UBERN’s expertise lies in the understanding and promotion of human behavior change. Using intensive longitudinal methods in daily life, their research investigates the patterns and mechanisms of human behavior as it relates to health. Based on these insights, they develop digital interventions to change behavior with the goal of health promotion and disease prevention.
Role within AIRIS
Within AIRIS, the AIHN team at UBERN contributes to several technical work packages, and leads Tasks 1.5 and 2.1 on AI-based data integration and mechanism-informed causal modelling, respectively.
In Task 1.5, UBERN develops AI agents that enable researchers to interact with AIRIS data and analytical capabilities through natural language. These agents support data linking, harmonisation, synthesis and analysis across heterogeneous datasets and connect to the AIRIS knowledge infrastructure to facilitate transparent exploration of data and evidence. UBERN additionally contributes expert-in-the-loop annotation infrastructure and expertise in multimodal representation learning.
In Task 2.1, UBERN formalises biological mechanisms for integration into AIRIS generative models and develops the AIRIS Mechanism Library, combining reusable causal-model templates, causal-discovery methods and mechanistic knowledge. The approach integrates established biological knowledge with data-driven causal discovery, rather than learning causal structures solely from data.
Within AIRIS, the team of the Department of Health Psychology and Behavioral Medicine mainly contributes to the topic of Ethical, Legal, and Societal Implications and Trustworthiness (WP6), and leads Task 6.5 on Trustworthiness and Robustness Safeguards. Therein, they address the behavioral impact of the AI-driven decision support on human decision makers, but also the influence of human behavior on model performance.