PHUSE has published a new Imaging Data Anonymisation Guideline, establishing a common approach for preparing clinical trial imaging data for responsible secondary use.

Developed through the PHUSE Data Transparency Working Group and led by Biogen, the guideline reflects input from imaging, privacy, and data-sharing experts at Vivli, Sanofi, Pfizer, Novartis, Roche, and Databricks. It addresses a challenge that has become increasingly important as researchers seek to reuse richer forms of clinical research data, including MRI and CT scans.

Making imaging data easier to share responsibly

Clinical imaging can provide information that is not captured in traditional trial datasets, but sharing it introduces distinct privacy considerations. Identifying information may appear in DICOM file headers, be embedded directly in an image, or, in the case of head scans, be recoverable from facial anatomy.

Until now, organizations have often had to develop their own approaches for addressing these risks. The new guideline provides a shared framework for doing so, helping bring greater consistency to how imaging data is prepared before it is made available for secondary research.

A practical standard for anonymisation

The guideline sets out minimum expectations for anonymising DICOM imaging data within secure, governed research environments. It covers the replacement of patient identifiers with non-reversible codes, date shifting that preserves clinically meaningful intervals, removal of identifying text and facial features, and documentation of the anonymisation process.

It also recognizes that anonymisation cannot eliminate re-identification risk entirely. For that reason, the guideline positions technical anonymisation alongside controlled access and governance as complementary safeguards.

Supporting more consistent data sharing

For organizations sharing clinical trial imaging data, the guideline offers a practical reference point developed and reviewed across industry, research, and data-sharing organizations. Greater consistency in how imaging data is prepared can reduce uncertainty for data contributors while supporting broader access to valuable data for secondary research.

Read the Imaging Data Anonymisation Guideline (Links to PDF download)