Systematic Redaction for Neuroimage Data

Systematic Redaction for Neuroimage Data

Matt Matlock (Center for Biological Systems Engineering, Department of Pathology and Immunology, Washington University School of Medicine, USA), Nakeisha Schimke (Institute of Bioinformatics and Computational Biology, Tandy School of Computer Science, University of Tulsa, USA), Liang Kong (Institute of Bioinformatics and Computational Biology, Tandy School of Computer Science, University of Tulsa, USA), Stephen Macke (Institute of Bioinformatics and Computational Biology, Tandy School of Computer Science, University of Tulsa, USA) and John Hale (Institute of Bioinformatics and Computational Biology, Tandy School of Computer Science, University of Tulsa, USA)
DOI: 10.4018/jcmam.2012040104


In neuroscience, collaboration and data sharing are undermined by concerns over the management of protected health information (PHI) and personal identifying information (PII) in neuroimage datasets. The HIPAA Privacy Rule mandates measures for the preservation of subject privacy in neuroimaging studies. Unfortunately for the researcher, the management of information privacy is a burdensome task. Wide scale data sharing of neuroimages is challenging for three primary reasons: (i) A dearth of tools to systematically expunge PHI/PII from neuroimage data sets, (ii) a facility for tracking patient identities in redacted datasets has not been produced, and (iii) a sanitization workflow remains conspicuously absent. This article describes the XNAT Redaction Toolkit - an integrated redaction workflow which extends a popular neuroimage data management toolkit to remove PHI/PII from neuroimages. Quickshear defacing is also presented as a complementary technique for deidentifying the image data itself. Together, these tools improve subject privacy through systematic removal of PII/PHI.
Article Preview


Data sharing facilitates collaboration and enables reproducibility, peer review, and meta-analysis studies. This is especially significant in neuroimaging studies, where subject enrollment tends to be low (Costafreda, 2009; Van Horn et al., 2009). Collaboration allows institutions to pool data to increase the number of scans and diversity of subjects, improving statistical power and reliability.

The notable success of the Alzheimer's Disease Neuroimaging Initiative (ADNI), a multisite neuroimaging project, demonstrates the potential for global inter-organizational collaboration (Mueller et al., 2005; Kolata, 2010; Toga et al., 2010). Data has been collected from 57 sites and images distributed to more than 1,300 investigators to date (Weiner et al., 2012). ADNI data has led to over 250 publications, and the effort has inspired similar data sharing initiatives for other conditions.

Collaborative efforts of this scale, however, introduce new challenges. Technical solutions to data storage, transmission, management, and dissemination problems continue to evolve through the development of medical imaging data management tools like the eXtensible Neuroimage Archiving Toolkit (XNAT), enabling a certain ease of data access (Marcus et al., 2007). However, none of these sharing solutions offer adequate and systematic tools for the preservation of subject privacy. Within these systems, subjects and their associated data must be managed by deidentifying the records before being added to the storage solution. While automated tools have been developed to remove metadata, none accomplish this deidentification on a systematic, study-wide (or institution-wide) basis (Neuroinformatics Research Group, 2010; MathWorks, 2012).

Any systematic tool for deidentification must be adaptable to the needs of the organization and research studies that it serves. Depending on situational research requirements, different aggregations of PHI/PII in publicly available datasets must be explicitly disallowed. The need for investigators to carefully consider the impact of their data is essential. Once a policy is decided upon, its implementation should involve minimal effort and integrate with ease into the data management workflow. In terms of its accessibility, popularity and open nature, XNAT presents itself as a viable platform for the integration of comprehensive privacy preservation processes for medical images.

Complete Article List

Search this Journal:
Volume 4: 2 Issues (2014)
Volume 3: 4 Issues (2012)
Volume 2: 4 Issues (2011)
Volume 1: 4 Issues (2010)
View Complete Journal Contents Listing