Project: GPT-MEDIC (ERC Starting Grant)
Host: EPICENTER, Medical University of Innsbruck
Supervisor: Patrick Rockenschaub
Start: as soon as possible (flexible)
Duration: 3–4 years, fully funded
In this ERC-funded PhD, you will develop methods for pre-training on ICU event streams across multiple hospitals. The core challenge is building foundation models that generalise across sites without learning site-specific shortcuts or memorising sensitive information.
ICU data are not like text or images. Events are irregular, values continuous, and measurements depend on clinical decisions that vary by hospital, clinician, and patient state. These quirks make generalisation difficult and privacy risks real, which is why your work will explore multicentre and potentially federated approaches to pre-training. Federated learning matters in particular because ICU data often cannot be pooled.
The goal is foundation models that transfer across settings and support many clinical outcomes. You will have access to ~1M ICU patients and ~33B events, with around half available from day one. The project involves close collaboration with Amsterdam UMC, UCL, and Cambridge.
The group
You will join a small, recently established research group “AI for Clinical Decision-Making” within a larger institute. You will have regular one-on-one supervision and direct access to clinical and technical collaborators across partner sites.
What we are looking for
* MSc in computer science, statistics, mathematics, data science, or related field
* Strong Python skills and experience with PyTorch, JAX, TensorFlow, or similar
* Demonstrated ability to complete a machine learning or data science project
* Experience with clinical data, federated learning, or data privacy is helpful but not required.
What we offer
* Fully funded position at € 39,005.40 gross/year (30 hours/week) per university collective agreement
* PhD program in Digital Medicine, with structured coursework and a peer cohort
* Shared access to ~30 H200 GPUs
* Conference travel and publication support
Women and underrepresented candidates are especially encouraged to apply. German is not required. You will be expected to relocate to Innsbruck, Austria.
#J-18808-Ljbffr