Enabling the sharing of neuroimaging data through well-defined intermediate levels of visibility

• Published in: NeuroImage (Orlando, Fla.) Vol. 22; no. 4; pp. 1646 - 1656
• Main Authors: Smith, Kenneth, Jajodia, Sushil, Swarup, Vipin, Hoyt, Jeffrey, Hamilton, Gail, Faatz, Donald, Cornett, Todd
• Format: Journal Article
• Language: English
• Published: United States Elsevier Inc 01.08.2004; Elsevier Limited
• Subjects: Academies and Institutes; Alzheimer Disease - diagnosis; Brain - pathology; Brain research; Computer Security; Diagnostic Imaging; Follow-Up Studies; Health Insurance Portability & Accountability Act 1996-US; Human subjects; Humans; Information Storage and Retrieval; Intellectual Property; Internet; Medical imaging; Metadata; Neuroimaging; Research Design; Responsibilities; Risk Management; Semantics; Sharing; Software; Neuroimaging; Database; Sharing
• ISSN: 1053-8119; 1095-9572
• DOI: 10.1016/j.neuroimage.2004.03.048

The sharing of neuroimagery data offers great benefits to science, however, data owners sharing their data face substantial custodial responsibilities, such as ensuring data sets are correctly interpreted in their new shared context, protecting the identity and privacy of human research participants, and safeguarding the understood order of use. Given choices of sharing widely or not at all, the result will often be no sharing, due to the inability of data owners to control their exposure to the risks associated with data sharing. In this context, data sharing is enabled by providing data owners with well-defined intermediate levels of data visibility, progressing incrementally toward public visibility. In this paper, we define a novel and general data sharing model, Structured Sharing Communities (SSC), meeting this requirement. Arbitrary visibility levels representing collaborative agreements, consortium memberships, research organizations, and other affiliations are structured into a policy space through explicit paths of permissible information flow. Operations enable users and applications to manage the visibility of data and enforce access permissions and restrictions. We show how a policy space can be implemented in realistic neuroinformatic architectures with acceptable assurance of correctness, and briefly describe an open source implementation effort.