The pilot project AI in Medicine, embedded in Come2Data, addresses data-related challenges in medical research and application contexts.

Its focus is on AI methods for standard medical data types, including electronic health records, pathology and radiology imaging data, as well as genomic and text-based information. The project covers both established machine learning and deep learning approaches, as well as predictive models and larger foundation models.

Medical data  pose highly specific challenges, because their access, processing, and reuse are subject to strict regulatory, ethical, and data protection requirements. The project therefore investigates how AI applications can be developed and applied responsibly, robustly, and effectively under these conditions.

In close interaction with Come2Data, the pilot project contributes to the development of best-practice guidelines and the identification of common data-related practices across domains.

It is also part of the help desk, supports knowledge transfer through workshops, and benefits from advisory and infrastructure services – in particular trusted research environments for the secure handling of sensitive medical data.