How the Medical Informatics Platform helps researchers securely analyse patient data

A new article in Frontiers in Disaster and Emergency Medicine explains how the Medical Informatics Platform (MIP), developed within EBRAINS and now part of the EBRAINS Research Infrastructure, enables researchers to analyse large clinical datasets while protecting patient privacy.

Hospitals and emergency departments collect large amounts of patient information in electronic health records. These data could be very useful for medical research and for improving patient care, but sharing them between hospitals and countries is difficult because of privacy laws, different data systems, and restrictions on moving patient-level data.

The article presents two real-world federations using the MIP: FERES, which connects stroke registries across Europe, and eCREAM, which focuses on emergency department data. Both show how the MIP can support cross-border research while allowing each institution to remain in control of its own data.

The MIP follows a “code-visits-data” model. Instead of sending patient records to a central database, each hospital or registry keeps its anonymized data on a secure local server called a node. When researchers run an analysis, the computation is sent to each node, and only aggregated results are returned to the central interface. The actual patient data never leave the institution where they are hosted.

Using a web-based interface, doctors and researchers from different institutions can work together to analyse large datasets from several hospitals in near real time. Data are harmonized and anonymized before analysis, access is controlled through governance procedures, and the platform applies privacy safeguards to prevent identification of individual patients.