Release
Date: August 26, 2011
Response Date: Received by October 14, 2011
NIH Blueprint for Neuroscience Research (http://neuroscienceblueprint.nih.gov)
National Institute of Mental Health (NIMH)
National Eye Institute (NEI)
National Center for Complementary and Alternative Medicine (NCCAM)
National Institute on Aging (NIA)
National Institute on Alcohol Abuse and Alcoholism (NIAAA)
National Institute of Biomedical Imaging and Bioengineering (NIBIB)
National Institute of Child Health and Human Development (NICHD)
National Institute on Deafness and Other Communication Disorders (NIDCD)
National Institute of Dental and Craniofacial Research (NIDCR)
National Institute on Drug Abuse (NIDA)
National Institute of Environmental Health Sciences (NIEHS)
National Institute of General Medical Sciences (NIGMS)
National Institute of Neurological Disorders and Stroke (NINDS)
National Institute of Nursing Research (NINR)
Office of Behavioral and Social Sciences Research (OBSSR)
This Request for Information (RFI) is for information and planning purposes only, and should not be construed as a solicitation or as an obligation on the part of the Federal Government, the National Institutes of Health (NIH), the NIH Blueprint for Neuroscience Research (Blueprint), and/or the National Institute on Drug Abuse (NIDA). The NIH does not intend to make any awards based on responses to this RFI or to otherwise pay for the preparation of any information submitted or for the Government's use of such information.
The Blueprint Workgroup on Neuromaging Data-Sharing and the Blueprint Lifespan Human Connectome Project Team are seeking input from the scientific community, health professionals, patient advocates, community-based organizations, students, and the general public about the administrative and technical issues involved in removing barriers to the sharing of human brain imaging data collected using NIH support. This input may include (but is not limited to) issues or potential solutions related to image hosting, as well as the adoption of standardized scanning procedures, imaging protocols, phenotypic characterization, and data formats. Of particular interest is input regarding human-subjects protection and confidentiality issues.
Definitions: Standardization, for the purpose of this Notice, is defined as the joint adoption (across multiple laboratories and institutions) of either: a) specific neuro-image acquisition parameters, b) image processing processes after image collection and reconstruction, c) specific instruments for collateral phenotypic characterization, d) data formats and definitions, e) data de-identification, or f) potential post-facto algorithms or schemes for harmonizing data collected under different parameters or data definitions. Sharing, for the purpose of this Notice, is defined as either the actual transfer of brain image data from the laboratory infrastructure of initial data collection or simply as enabling dynamic access to image data that remains stored at the institution of its collection. In either case, image data would be annotated with demographic and other phenotypic information in sufficient detail for secondary data analysis. Transferred data could be sent directly to a particular laboratory (either through a centralized referral hub, or bilaterally on request), or alternatively furnished to one or more common repository infrastructures.
Background
The NIH Blueprint for Neuroscience Research endorses the principle of standardization and open-access to brain imaging data. For example, the Blueprint-funded Human Connectome Project (HCP) promises to provide the neuroscience community easy access to well-annotated raw and processed data for secondary analyses and potential federation with other datasets. Alternatively, the 1000 Functional Connectomes Project (FC1000) is a consortium-based dataset comprised primarily of resting-state functional connectivity data obtained from healthy human subjects in dozens of individual laboratories (Biswal et al. Proc Natl Acad Sci (2010);107(10):4734-9). This consortium provided evidence that federated brain image datasets can successfully be compiled for secondary analysis, especially for exploratory investigations. The developing distribution/sharing infrastructure of the HCP, as well as success of the FC1000, has fostered not only other nascent grass-roots efforts to pool donated image data for secondary analysis, but has also increased interest of the NIH Blueprint in fostering the expansion of open-access to neuroimage data from NIH-funded grants, contracts, and cooperative agreements, where standardization across labs would significantly facilitate the ease of federation.
Expanded sharing of image data would be further enhanced by increased standardization across laboratories or repositories, and holds the potential for significant value-added investigations of the human brain at relatively minimal expense. For example, assembly of data across different research projects may allow investigations of genetic or behavioral associations that require sample sizes and statistical power too large for any one laboratory to collect. Alternatively, assembly of image data collected across multiple laboratories could enable investigations comparing brain metrics across disease states that span the programmatic foci of different NIH institutes and centers. Despite this potential, however, few laboratories or institutions currently provide easy access to stored neuroimage and related meta-data.
Information Requested
The NIH Blueprint is interested in removing barriers to sharing in order to expand access to NIH-funded neuroimage data for secondary data analysis or meta-analysis. This RFI seeks information that will help the NIH Blueprint: a) identify the interest, utility, preferences, and potential technical and administrative challenges in overcoming barriers to data-sharing, b) expand the standardization of brain image data, and c) facilitate the sharing of brain image data. Individual members of the scientific community, scientific organizations, healthcare professionals, patient advocates, and the public are invited to provide commentary on the following items:
1. Describe the current barriers to the standardization and sharing of neuroimaging data for secondary analysis. Responses can include comments or suggestions about strategies to incentivize data sharing.
a. Mandatory (condition of award or mandated by journals)
b. Incentives to promote voluntary sharing and adoption of standard procedures
c. Technical or infrastructure/IT barriers
2. Considerations of human-subjects protections
a. Issues and experiences related to informed consent of subjects for their data to be:
- stored indefinitely
- stored in a repository other than of the laboratory of data collection
- available to third parties for secondary analysis
b. Issues related to subject confidentiality, with respect to
- de-identification of brain image data
- potential for identification of subjects based on brain images (such as gyrus and sulcus morphology)
- extent of access for secondary analysis
- HIPAA compliance
3. Suitability of different image data modalities for donation and federation across labs, and utility of different image modalities for secondary analysis
a. Resting-state or task-evoked functional MRI
b. Structural MRI (e.g. diffusion or volumetric)
c. Other image modalities like MRS or PET
d. Reconciling the need for standardization versus the value of methods innovation
4. Conditions or qualifications under which data should be shared
a. Applicant eligibility or qualifications; composition of data access committee
b. Data quality or sample size
c. Embargo by collecting lab (duration) before furnishing data
d. Financial considerations (e.g. fee-for-access, current NIH funding status)
5. Existing (including international) efforts, large scale databases, clearing houses, tools, and resources that facilitate sharing of imaging data
a. Lessons learned
b. Features and suitability for up-scaling to accommodate more data- especially of different modalities, or with different linked phenotypic data
c. Methods for harmonizing data from different extant or emerging extramural repositories
6. Issues in standardization, sharing, and maintenance of image data and linked phenotypic or genetic data
a. Advantages, disadvantages, and costs of centralized repositories versus distributed hosting
b. Computational infrastructure, database management, and data processing software/algorithms
c. Data and meta-data formats, common data elements
d. Challenges, costs, sources, and other issues of long-term data storage (curation)
e. Linking image data with genetic or genomic data
f. Interface with centralized referral hubs, the HCP, or similar NIH-initiated infrastructures
7. Issues of phenotypic harmonization or definitions with respect to:
a. Developmentally-sensitive measures (suitability at different life stages)
b. Objective behavioral metrics
c. Psychometric measures
d. Psychiatric disorders, other brain disorders, or histories of substance use
e. Family history, environmental or other psychosocial factors
f. Expansion of use (or a mandate) to adopt consensus phenotypic measures
The NIH Blueprint also welcomes any additional comments and suggestions with regard to considerations not mentioned above.
Note: Responses to NOT-DA-11-008 Request for Information (RFI): Creation Of A Consortium-Based Repository of fMRI-Based Connectivity Brain Scans From Substance-Using or Abusing Clinical Research Participants will be collated with responses to this RFI.
Responses will be accepted until October 14, 2011 via email to: James Bjork, Ph.D. at jbjork@mail.nih.gov. Please mark your responses with this RFI identifier NOT-DA-11-021. Reponses are expected to be no longer than approximately 2000 words.
Respondents will receive an automated email confirmation acknowledging receipt of their response, but will not receive individualized feedback.
Any identifiers (e.g., names, institutions, e-mail addresses, etc) will be removed when responses are compiled. Only the processed, anonymized results will be shared internally with the NIH Blueprint program staff and leadership, as appropriate. Nonetheless, no proprietary information should be submitted.
Please direct all inquiries to:
James M. Bjork, Ph.D.
Clinical Neuroscience Branch
Division of Clinical Neuroscience and Behavioral Research
National Institute on Drug Abuse
National Institutes of Health
6001 Executive Boulevard, Room 3172
Bethesda, MD 20892
Phone: (301) 443-3209
Email: jbjork@mail.nih.gov
and Human Services (HHS)
