MRI Data Protection

This slide deck was used in the afternoon workshop at the Data Protection Day 2026, which took place on 22 June 2026 at the University of Zurich. Summary of the workshop Researchers are expected to share data promptly and “as open as possible, as closed as necessary”. Finding the right balance between these two concepts has become increasingly difficult with recent improvements in artificial intelligence and increasing availability thereof. Fairly recent standard de-identification approaches, such as defacing, can no longer be considered safe in view of performance refacing algorithms developed in the last decade. This workshop aims to give participants an overview of: current best practices and privacy protective measures overview of some available tools and concepts to achieve best possible de-identification a workflow example allowing researchers to publish their data FAIR, especially when open data is not possible