JADS project develops new AI technology protecting patients’ privacy


AI has proven to be very effective in the diagnosis of illnesses and digital medicine pave the way for new opportunities in healthcare. At the same time, these algorithms detecting and identifying patterns for the diagnosis learn from very sensitive and personal data, namely medical data.

Our JADS project “representation learning in oncologic imaging” deals with this important and crucial topic. Together with the non-profit organization OpenMined, the team has develop a new AI technology for radiological image data that safeguard data privacy in the training of algorithms. We congratulate the research team with researchers from both TUM and the Imperial College London on their first milestone and success of the project. PD Dr. Georgios Kaissis, Prof. Marcus Makowski together with Prof. Daniel Rueckert and PD Dr. Rickmer Braren have primarily contributed to this result by combining their expertise in the areas of medicine and artificial intelligence.

Further information and background on this medical advancement can be found in TUM’s research news.

The project is part of the first JADS cohort dealing with the topic of “AI, Robotics, and Healthcare” that has started in October 2020. Both TUM and Imperial, follow the objective with this program to support and foster cutting-edge research in trending and innovative research fields. This great result has shown that interdisciplinary research across departments and institutions truly brings great outcome. We are looking forward to further results from all our research teams. Learn more about JADS at our program page