Computational Medical Imaging

Medical foundation models, multimodal imaging and uncertainty — methods that turn images into reliable clinical evidence.

We develop the core methods of computational medical imaging: medical foundation and vision-language models, multimodal learning that fuses imaging with clinical data, and principled uncertainty estimation so that models know when they do not know. Our work spans radiology (MRI, CT, X-ray, ultrasound, mammography), histopathology, ophthalmology and endoscopy.

Key questions

  • How do we build foundation models that transfer across organs, modalities and tasks with little labelled data?
  • How can imaging, text and clinical data be combined into one coherent model?
  • How do we quantify and communicate uncertainty to clinicians?

Funded projects 33 filter all →

Publications 155 filter all →

KI-gestützte datenschutzkonforme multizentrische Vorhersage lymphogener Metastasierung beim Pankreaskarzinom mittels eines föderierter CT-Foundation-Modelle

Maximilian Coerper, Parinishtha Bhalla, David D. Gaviria, Patrick Kupczyk, Lena‐Christin Conradi, Ali Seif Amir Hosseini, D Ma, S Jacob, Ioannis Pozios, Matthäus Felsenstein, Tim Oliver Vilz, Shadi Albarqouni, Alexander Semaan

Zeitschrift für Gastroenterologie, 2026

Journal Medical Imaging

All 155 publications →

Datasets 11

Demos 4

Code 11

Teaching 14

In the media 2

Community & events 68