ÑÇÖÞÌìÌø£Àû

In the course of the digital revolution, the use and processing of personal data has become important for the overall value chain. In this context, it is vital to ensure a balance of interests among all stakeholders. While business and research are interested in a smooth exchange of data, the people providing data want above all to preserve their data sovereignty in terms of data protection law and data ethics. One key to reducing the threat of asymmetries lies in the involvement of data trustees, i.e. neutral intermediaries.

In order to establish data trustees in the long term, a legal regulatory framework is needed that guarantees data protection and at the same time creates incentives to promote innovation. TreuMed is taking up this challenge. We develop and test data trustee models and corresponding business concepts using the example of "distributed artificial intelligence" in medicine.

Machine learning methods in medical research rely on extensive, sensitive patient data to be able to make reliable predictions. However, strict data protection can counteract the success of research because it reduces the amount of data available. This in turn has a negative impact on the success of patient treatment. TreuMed provides a technical solution to this dilemma with "distributed artificial intelligence" in medicine. The core of the TreuMed concept is a traffic light system for data trustees: depending on the identifiability of the patient data, one of three possible privacy levels and a corresponding level of protection is determined. The model is flanked by certification obligations and liability concepts for residual risks. Furthermore, the presented models are tested through applications of "distributed artificial intelligence" in the field of molecular epidemiology and biomarker research using the example of (identifying) genomics and transcriptomics data.