1 of 24

​

​

2 of 24

Open Access

Open Data

Research Integrity

Transparency

Open Methodology

Open Scientific workflows

Open Science

3 of 24

4 of 24

Why OSCL?

  • Learn from peers and from other disciplines; your colleague down the hallway may have hands-on experience, your colleague from another faculty may tell you how their field solved an issue.
  • Discuss and stay updated with recent advances; There are a lot of new terms and keeping up is both necessary and time-consuming.
    • Mentor others; there are always people who are newer to things than you are—help them.

​

5 of 24

Where OSCL?

6 of 24

7 of 24

How OSCL?

https://www.universiteitleiden.nl/open-science-community-leiden/join-us

8 of 24

Sharing & archiving MRI data

Today: Hackathon

9 of 24

Aims of today

  • identify problems

​

  • identify what we need to solve problems

​

  • where to go from here

​

  • (if time: start formulating solutions)

​

10 of 24

Who are you?

11 of 24

Program

  • Some background:
    • Privacy (GDPR / AVG)
    • Take of: Sander Nieuwenhuis
    • Platforms (Roelof van der Kleij)
  • Brainstorm: main issues
  • Break
  • Brainstorm: what needed to solve issues?
  • If time: Brainstorm: formulating solutions

12 of 24

Privacy (GDPR)

  • Personal data: can be traced back to a person
    • Process & share: base on lawful ground, e.g.,:
      • Informed consent (prior to start research)
      • “Public interest”
    • Pseudonymous = still personal data

​

  • Anonymous data: cannot be traced back to a person → no personal data
    • Process & share: allowed

​

13 of 24

Privacy & MRI data

  • Raw MRI data: personal

​

  • Deidentification:
    • file names
    • header files

(names, acquisition dates)

    • defacing tools

​

  • Deidentified ≠ anonymous?
    • Anonymous to whom? Researcher/user/both?
    • Now vs. future (e.g., machine-learning)

​

14 of 24

Don’t be afraid; keep thinking!

​

​

​

​

15 of 24

Platforms

​

​

  • Publication packages (mandatory):

all that belongs to a publication

  • ✅ Persistent identifier; long-term

storage; public/on request

  • ❌ Data limit

​

16 of 24

  • Project management / workflow
  • ✅ Persistent identifier; long-term storage; public/by request/private
  • ❌ 5 GBs per file

​

​

​

17 of 24

​

​

  • Raw (deidentified) MRI data storage and sharing
  • ✅ Persistent identifier; long-term

storage; no size limit; machine-readable (BIDS format)

  • ❌ Always public after 3 years, USA-based

​

18 of 24

  • Unthresholded statistical maps, parcellations, and atlases of the brain | Can belong to a publication
  • ✅ Persistent identifier; long-term storage; no size limit; allows meta-analyses; anonymous data
  • ❌ Only group-based data (e.g. contrasts)

​

​

​

19 of 24

​

  • OMERO

​

  • Roelof van der Kleij, ISSC

​

​

​

​

20 of 24

Brainstorm 1

​

Which issues would/do/can you run into when attempting to share and/or archive MRI data?

​

  • Small groups, 4-6

​

  • E.g.,:
    • No knowledge/information/experience
    • No storage space (platforms)
    • No consent asked (privacy)
    • Which data to share? (e.g., student vs. patient, raw, preprocessed, first level, second level)
    • How to connect to other research output?
    • … etc.

21 of 24

Break

    • ​

22 of 24

Brainstorm 2

What do you need to be able to solve these issues?

​

  • Small groups, 4-6

​

  • E.g.,:
    • More information
    • Someone to help
    • Storage space
    • Better informed consent forms
    • … etc.

23 of 24

Brainstorm 3

(only if time)

Formulating solutions

​

  • Small groups, 4-6

​

  • E.g.,:
    • Draft IC forms
    • Work out step-by-step plan: how to?
    • Work out an example case
    • Choose best suitable platform at the moment
    • … etc.

24 of 24

Thank you

​

​