K-anonymity framework for robust health data sharing published in IJIS
A joint privacy-engineering effort showing how health datasets can be shared for research while resisting re-identification — published in the International Journal of Information Security.

Mastering data privacy: leveraging K-anonymity for robust health data sharing, by Stylianos Karagiannis, Christoforos Ntantogian, Emmanouil Magkos, Aggeliki Tsohou and Luís Landeiro Ribeiro, tackles one of the most stubborn tensions in applied privacy: health data is enormously valuable for research and enormously dangerous to release.
The work examines how K-anonymity and related generalisation models can be applied to real health datasets so that records remain useful for analysis while resisting re-identification through quasi-identifiers and linkage attacks. It is characteristic of the lab's approach — a formal privacy model, evaluated against practical utility requirements rather than in the abstract.
Connected work
- Chidroid — a mobile Android application for log collection and security analysis in healthcare and IoMT (Applied Sciences, 2023).
- Cybersecurity and Medical Imaging — a simulation-based approach to DICOM communication (Applied Sciences, 2023).
- The anonymisation layer for trainee data in the NITRO cyber range, which applies the same thinking to training telemetry.
Together these form a healthcare security and privacy strand that runs across both the cybersecurity and privacy research fields and the lab's collaboration with PDMFC in Portugal.