Fig. 1

Software Architecture. The data layer comprises two relational databases to store patient identifying data and pseudonymized payload data separately. The business layer adds an application server with a pseudonymization service, a visualization server, as well as a server supporting model simulations (MAGPIE). In particular, the application server provides the access to patient identifying data and pseudonymized payload data (1). The visualization server is strictly separated from the identifying patient database and exclusively retrieves medical data from the pseudonymized payload database for data description and model prediction (2 and 3). The presentation layer provides the frontend with a web-based graphical user interface for onsite access by physicians. The php- and R-logo are taken from the websites http://php.net/download-logos.php and https://www.r-project.org/logo/. Both images are under the terms of the Creative Commons and Attribution-Share Alike 4.0 International (CC-BY-SA 4.0)