Commit b13ca243 authored by Cassandra Gould van Praag's avatar Cassandra Gould van Praag
Browse files

add docs

parent 85549e8c
---
title: 'Open Data Process Diagram - Appendices'
---
Appendix 1 - DICOM Headers to be scrubbed
=========================================
List of DICOM identifying fields that need to be scrubbed included in
McGill Centre for Integrative Neuroscience Open Science Guidance Open
Science Data Preparation Checklist.
Fields to be scrubbed include anything which is identifiable as
participant information or unique to the scanning session or data
acquisition site. Fields identified by McGill have been checked against
DICOM fields appearing in data acquired on WIN 3T (OHBA) for human
participant.
----------------------------------- ----------------------------------------------------------------------------------------------------------------------------------
**Field name** **In WIN dicom example (if field present)**
PatientAddress n/a
PatientAge 042Y
PatientBirthDate+D10 19751005
PatientBirthName n/a
PatientID W3T\_2016\_11\_090
PatientMotherBirthName n/a
PatientName W3T\_2016\_11\_090
PatientReligiousPreference n/a
PatientSex F
PatientTelephoneNumbers n/a
Patient Size 1.62
Patient Weight 64
OtherPatientIDs n/a
OtherPatientName n/a
OtherPatientNames n/a
AcquisitionDate 20180515
AcquisitionTime 90931.46
ContentDate 20180515
ContentTime 384 values: 091507.445000, 091507.448000, 091507.450000, 091507.452000\...
DeviceSerialNumber 66093
FrameOfReferenceUID 1.3.12.2.1107.5.2.43.66093.1.20180515090420048.0.0.0
InstanceCreationDate 20180515
InstanceCreationTime 384 values: 091507.445000, 091507.448000, 091507.450000, 091507.452000\...
InstitutionAddress Warneford Lane StreetNo,Oxford,District,GB,OX3 7JX
InstitutionalDepartmentName BE3597
Institution Name Warneford Hospital
IssuerOfPatientID n/a
MediaStorageSOPInstanceUID n/a
OperatorsName n/a
PerformedProcedureStepDescription n/a
PerformedProcedureStepID 0
PerformedProcedureStepStartDate 20180515
PerformedProcedureStepStartTime 85830.014
PerformingPhysicianName Blank
PhysiciansOfRecord n/a
ReferencedSOPInstanceUID n/a
ReferringPhysicianName Blank
RequestedProcedureDescription n/a
RequestingPhysician Blank
SeriesDate 20180515
SeriesInstanceUID 2 values: 1.3.12.2.1107.5.2.43.66093.2018051509150468746204326.0.0.0, 1.3.12.2.1107.5.2.43.66093.2018051509093178111004205.0.0.0
SeriesTime 2 values: 091507.443000, 091508.436000
SOPInstanceUID 384 values: 1.3.12.2.1107.5.2.43.66093.2018051509150469069804346, ...
StationName AWP66093
StudyDate 20180515
StudyDescription OHBA Projects\^2018\_108 RESTAND
StudyInstanceUID MR20180515085649
StudyTime 85829.922
----------------------------------- ----------------------------------------------------------------------------------------------------------------------------------
Appendix 2 - BIDS Compliance
============================
Data should be confirmed as BIDS compliant using the BIDS-validator
docker image available from
[[https://github.com/bids-standard/bids-validator\#docker-image]{.underline}](https://github.com/bids-standard/bids-validator#docker-image).
This could be installed on XNAT.
Appendix 3 - Scrubbing .json files
==================================
.json sidecar files are generated for each nifti image to conform with
BIDS specification. These files contain essential metadata which
describes the acquisition of the data. They may also contain potentially
identifiable information relating to the participant which should not be
distributed openly.
Appendix 4 - Quality Control
============================
Quality control (QC) assessments should be completed as a matter of
course for research data, and as a courtesy for those wishing to reuse
your data. QC can be performed on BIDS formatted data using mriqc. Mriqc
is available as a container on XNAT.
Appendix 5 - Defacing
=====================
Consider using fsl\_deface. Other defacing tools are available.
Consider running QC on defaced images using
[[https://github.com/raamana/visualqc/blob/master/docs/VisualQC\_TrainingManual\_v1p4.pdf]{.underline}](https://github.com/raamana/visualqc/blob/master/docs/VisualQC_TrainingManual_v1p4.pdf)
Appendix 6 -- Upload to XNAT
============================
TBD (not exhaustive)
1. Default access groups (e.g. uploading researcher, PI and XNAT Admin)
2. Verification and management of external account (whole process to be
defined for external users) -- Look for external (acceptable) models
of this.
3. Process for verifying that the mandatory procedures have been
completed (e.g. DICOMS scrubbed) before data are made publicly
available.
Appendix 7 -- Data freeze
=========================
The data freeze process as applied for the Dementias Platform UK (DPUK)
is computationally expensive -- multiple freeze generates duplicates of
data. Alternative methods could be investigated (e.g. ZFS snapshots).
Consider developing a cost model of the platform to indicate to
researchers the cost of snapshots carried by WIN. A cost model would
also be useful for project and infrastructure grant applications.
Appendix 8 - Digital object identifiers (DOI)
=============================================
A DOI is essential to ensure that the data shared can be effectively
attributed to the authors via [[FAIR
principles]{.underline}](https://www.go-fair.org/fair-principles/).
Oxford Research Online (ORA) is the University's preference for
generating a DOI.
The ORA entry will link to an XNAT URL for the project (and/or
appropriate freeze). Major updates to XNAT (likely over the medium to
long term) will potentially lead to changes in URL, thus we cannot
guarantee they are persistent, as required for a DOI.
Routes considered to address URL persistence:
1. We create and manage URL redirects if/when changes to XNAT (or the
entire server becomes redundant)
2. ORA amend DOI entries when they are informed of a change or if they
are running periodic checks.
Research Data (<https://researchdata.ox.ac.uk>) agree that ORA will be
motivated to carry some of the burden, with their focus on effective
curation. Opened discussions with ORA on 10^th^ June 2021.
Appendix 9 - Data usage agreement (DUA)
=======================================
Research Services (<research.services@admin.ox.ac.uk>.) can provide
template agreements for data sharing.
A range of template data sharing and data access agreements (suitable
for research from all four academic divisions) are available. While
these agreements are designed to address key issues, they will need to
be tailored to the specific needs of a particular project. Research
Services can advise on this, and also offer to check drafts as they are
prepared before producing a final signature-ready copy.
Example DUA from Donders Repository ([https://data.donders.ru.nl/](https://data.donders.ru.nl/?0))
--------------------------------------------------------------------------------------------------
I request access to the data collected in the digital repository of the
\<DEPARTMENT\>, part of the \<INSTITUTION\>, established at \<CITY\>,
\<COUNTRY\> (hereinafter referred to as the \<INSTITUTION SHORTNAME\>).
By accepting this agreement, I become the data controller (as defined
under the GDPR) of the data that I have access to, and am responsible
that I access these data under the following terms:
1. I will comply with all relevant rules and regulations imposed by my
institution and my government. This agreement never has prevalence
over existing general data protection regulations that are
applicable in my country.
2. I will not attempt to establish or retrieve the identity of the
study participants. I will not link these data to any other database
in a way that could provide identifying information. I shall not
request the pseudonymisation key that would link these data to an
individual's personal information, nor will I accept any additional
information about individual participants under this Data Use
Agreement.
3. I will not redistribute these data or share access to these data
with others, unless they have independently applied and been granted
access to these data, i.e., signed this Data Use Agreement. This
includes individuals in my institution.
4. \[OPTIONAL\] When sharing secondary or derivative data (e.g. group
statistical maps or templates), I will only do so if they are on a
group level, and cannot be deduced information from individual
participants.
5. I will reference the specific source of the accessed data when
publicly presenting any results or algorithms that benefited from
their use:
a. Papers, book chapters, books, posters, oral presentations, and
all other presentations of results derived from the data should
acknowledge the origin of the data as follows: "Data were
provided (in part) by \<Research centre/University Department\>
\<University, Country\>".
b. Authors of publications or presentations using the data should
cite relevant publications describing the methods developed and
used by the \<Research centre/University Department\> to acquire
and process the data. The specific publications that are
appropriate to cite in any given study will depend on what the
data were used and for what purposes. When applicable, a list of
publications will be included in the collection.
c. Neither the \<Research centre/University Department\> or
\<University\>, nor the researchers that provide this data will
be liable for any results and/or derived data. They shall not be
included as an author of publications or presentations without
consent.
6. Failure to abide by these guidelines will result in termination of
my privileges to access these data.
Appendix 10 -- Advice from Research Data Oxford
===============================================
- Define what a breach looks like (how, why, impact) and demonstrate
that you have shown due diligence in preventing it.
- How will you manage accusations of disclosure?
- The consent you have is the cornerstone of your agreement with the
participant. Make sure consent appropriately describes all relevant
procedures including deidentification and managed access.
- Agree that (brain) MRI data cannot be considered anonymous and
therefore falls under GDPR.
- The DUA should preclude commercial use.
- Access to data should be monitored and reported on (for
accountability and auditing)
Research Data would be pleased to partner with us on this project. They
are interested in growing their experience of working directly with
specific examples.
Acknowledgements
================
Created using materials from: Das et al, 2019, \"MCIN Open Science
Guidance: Data Preparation Checklist\"
[[https://loris.ca/MCINOpenScienceGuidance\_DataPrepChecklist.pdf]{.underline}](https://loris.ca/MCINOpenScienceGuidance_DataPrepChecklist.pdf)
Inspired by Huijser, Dorien, Achterberg, Michelle, Wierenga, Lara, Van
\'t Veer, Anna, Klapwijk, Eduard, Van Erkel, Raymond, & Hettne,
Kristina. (2020, June 19). MRI data sharing guide. Zenodo.
http://doi.org/10.5281/zenodo.3822290
Relevant policy and guidance
============================
University of Oxford Policy on the Management of Data Supporting
Research Outputs
<https://researchdata.ox.ac.uk/university-of-oxford-policy-on-the-management-of-data-supporting-research-outputs/>
3.0 Responsibilities of the researcher
3.1 Principal Investigators hold day-to-day responsibility for the
effective management of research data generated within or obtained from
their research, including by their research groups. This shall include
understanding and complying with the requirements of any relevant
contract with or grant to the University that includes provisions
regarding the ownership, preservation and dissemination of research
data.
3.2 Researchers will protect confidential, personal and sensitive
personal research data in accordance with legal and ethical requirements
related to the research they conduct.
3.3 Researchers will make every reasonable effort to keep an accurate
and comprehensive record of their research, including documenting clear
procedures for the collection, storage, use, reuse, access and retention
or deletion of the research data associated with their records.
3.6 Researchers should strongly consider depositing their data
supporting outputs in an appropriate data repository along with
sufficient descriptive metadata (a data record) to ensure that it can be
found and understood. Where data is deposited somewhere other than the
University's institutional data repository (the Oxford Research Archive
for Data, or ORA-Data), a data record should also be created in ORA-Data
which describes and points to the data.
Markdown is supported
0% or .
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!
Please register or to comment