Related Experiment Video
Updated: Jun 20, 2026
06:18
Pioneering Patient-Specific Approaches for Precision Surgery Using Imaging and Virtual Reality
Published on: April 5, 2024
Reversible anonymization of DICOM images using automatically generated policies
Michael Onken1, Jörg Riesmeier, Marcel Engel
1OFFIS - Institute for Information Technology, 26121 Oldenburg, Germany. onken@offis.de
Studies in Health Technology and Informatics
|September 12, 2009
Summary
This study introduces reversible anonymization for medical imaging data, separating identifying (IDATA) and non-identifying (MDATA) information. This method ensures data privacy while allowing full access when needed, improving diagnostic integrity.
Area of Science:
- Medical Imaging
- Data Privacy
- DICOM Standards
Background:
- Medical imaging applications, especially those with internet-based data access, require robust systems for separating identifying (IDATA) and non-identifying (MDATA) patient data.
- Current DICOM image formats often intermix IDATA and MDATA within a single object, posing challenges for data management and access control.
- Role-based access systems are crucial for controlling patient data organization and access in medical imaging databases.
Purpose of the Study:
- To develop and evaluate a method for reversible anonymization of DICOM objects, separating IDATA from MDATA.
- To enable dynamic reconstruction of original images by re-linking anonymized data using unique tokens.
- To create a framework for automatic generation and execution of anonymization policies based on DICOM standards.
Main Methods:
- Reversible anonymization of DICOM objects by substituting IDATA with unique anonymous tokens.
- Automatic generation of anonymization policies derived from DICOM standard text.
- Implementation of a framework utilizing the DICOM toolkit (DCMTK) for policy execution.
- Evaluation of the system using real-world medical images from the SKELNET project.
Main Results:
- Successful separation and reversible anonymization of IDATA and MDATA in DICOM images.
- Demonstrated ability to dynamically re-link anonymized data to reconstruct original images for authenticated users.
- The developed framework effectively supports various DICOM image types.
- The system performed efficiently and effectively on real-world test images.
Conclusions:
- The proposed approach offers a reliable method for reversible anonymization in medical imaging, crucial for data privacy and security.
- The automatic policy generation and DCMTK-based framework provide a versatile solution applicable to diverse DICOM-based projects.
- This technology significantly enhances the quality and integrity of diagnostics in image-focused medical fields.