Towards Deep Medical Imaging Biobanks – PHRT


Towards Deep Medical Imaging Biobanks

Short Summary

Data-driven approaches for patient care are rapidly gaining importance in healthcare. Machine learning and artificial intelligence methods are yielding promising technologies to improve efficiency and accuracy of patient care. The most crucial ingredient of these new technologies is large-scale databases. Through determining the relevant statistical relationships in existing data, machine learning and AI technologies can solve complex problems and help determining diagnosis and choose optimal treatment. This project aims to build the necessary infrastructure in Switzerland for curating large datasets that will fuel next generation machine learning and AI technologies for leveraging information in medical images, such as Magnetic Resonance Imaging (MRI).


This project will develop the infrastructure in university hospitals in Zürich, Basel and Bern, to allow curating large-scale extensive databases. These databases will enable researchers in Switzerland to tackle a large set of research problems and develop cutting edge technologies for interpreting and analyzing medical images.


The technologies developed here will be instrumental in enabling cutting edge research in AI and machine learning for improving the use of medical images for patient care. They will also provide an advantage to researchers in Switzerland and eventually Swiss technology and software industry related to AI, machine learning and healthcare.


In the last 5 years, large-scale databases of medical images have enabled developing crucial AI technologies that interpret images automatically and help clinicians in making diagnosis and planning a treatment course. However, existing databases are limited in their capacity. They do not contain data measured by the acquisition devices but only images, which are in reality a processed output of the machine. This limits the extend of the research and developments that can be achieved using these databases. For example, it becomes very difficult to develop tools that will improve the acquisition itself so that better images can be acquired for each patient.


Prof. Dr. Ender Konukoglu

Biomedical Image Computing, ETH Zürich


  • Prof. Sebastian Kozerke, ETHZ
  • Prof. Stefanie Krämer, ETHZ
  • Prof. Michaël Unser, EPFL
  • Dr. Bram Stieltjes, USB
  • Prof. Hatem Alkadhi, USZ
  • Prof. Matthias Guckenberger, USZ
  • Prof. Robert Manka, USZ
  • Dr. Valerie Treyer, USZ
  • Dr. Sebastian Winklhofer, USZ


In Progress

Funded by