Machine learning methods (so-called AI techniques), medical data and the challenges of preserving the anonymity of the patients behind these data volumes were the topics at the kick-off meeting of the roughly 40 experts who came together for the first time today in the Audimax on the Lübeck campus. Minister Dirk Schrödter, head of the State Chancellery of Schleswig-Holstein, also travelled from Kiel for the event. AnoMed is one of five competence clusters for anonymisation nationwide and is funded by the BMBF as a joint project with a total of 10 million euros. Around 5.5 million euros of the funding go directly to research at the University of Lübeck.
Prof. Dr. Gabriele Gillessen-Kaesbach, President of the University of Lübeck, opened the kick-off meeting of the AnoMed cluster in Lübeck: “In recent years we have built up expertise in IT security, medical informatics and machine learning. The Federal Government's trust in the AnoMed competence centre shows that this strategy has paid off.”
AnoMed aims to show medical users the risks of deanonymisation and the possibilities of the latest anonymisation techniques. To this end, novel anonymisation methods are being developed and a benchmark and test platform with a range of medical application scenarios is being built, on which stakeholders worldwide can interact with medical data and the latest results of anonymisation research, i.e. both protection mechanisms and attacks. Channelling anonymisation research towards medical applications is important and requires new approaches, believes Prof. Dr. Esfandiar Mohammadi from the Privacy & Security working group at the Institute of IT Security at the University of Lübeck and scientific director of the AnoMed cluster: “Progress in IT security and machine learning gives us new possibilities for strong anonymisation solutions to protect personal data. This opens up new fields of application in digitalised and personalised nutrition, medicine and healthcare.”
Within AnoMed, primarily publicly available data sets will be used to demonstrate risks. In this way, users can better assess in which application contexts these methods are still sufficient and in which situations more reliable methods and definitions are required. These considerations are also important for the state data strategy in Schleswig-Holstein. Digitalisation minister Dirk Schrödter said: “We have a firm goal in mind: to significantly improve data availability. AI applications in particular depend on it, and health data are an extremely valuable asset for research, clinical practice and development in medical technology. We must no longer leave data unused in data silos. If the personal data behind it is well protected, even more people will make their health data available because trust grows. More data ultimately means more possibilities to improve and individualise therapies. That is why we want research and development in our state to be at the cutting edge when it comes to data security and anonymisation. The competence centre for anonymity is the perfect institution for this.” The minister welcomed the close cooperation between academia and the Independent State Institute for Data Protection in this project.
The aim of the cluster is to design novel attacks and initial anonymisation solutions and to examine legal questions of anonymisation. It also ensures that the potential for innovation and new treatment methods enabled by reliable anonymisation techniques is brought into medicine and the medical technology industry, thus promoting health and economic growth in Germany and the EEA. Lübeck offers a locational advantage thanks to its existing networking of research in the fields of medicine, AI and medical technology. “The availability of large data volumes is of essential importance for the use of AI in medicine. The possibility of completely anonymising data is what creates the basis for broad application. Accordingly, AnoMed is a decisive piece of the puzzle in the AI strategy of the University of Lübeck,” explains Prof. Dr. Stefan Fischer, Vice President for Digital Affairs and Transfer at the University of Lübeck.
The consortium for this novel basic research project consists of the Independent State Centre for Data Protection, EppData GmbH, Perfood GmbH, UniTransferKlinik Lübeck GmbH, the German Research Center for Artificial Intelligence, the University of Hamburg, the University Medical Center Schleswig-Holstein, Fraunhofer IMTE and the University of Lübeck, which leads the project.
Contact for questions:
Prof. Dr. rer. nat. Esfandiar Mohammadi
scientific director of the competence cluster
University of Lübeck
Ratzeburger Allee 160
23562 Lübeck
Mail: esfandiar.mohammadi@uni-luebeck.de
Tel: +49 451 3101 6609
