Steering Group minutes 2023/01/19
Date: Thursday, January 19th, 2023
Topic | Relevant Links |
---|---|
Chris Rorden |
https://github.com/bids-standard/bids-specification/issues/1350 Updating https://bids.neuroimaging.io/bep_guide.html → https://github.com/bids-standard/bids-website/issues/269 |
Recognizing BIDS validated datasets
|
BIDS is not yet required in journals, we could reach out to journals for this. OpenNeuro is typically where BIDS datasets are shared. Action: add journal level policy for BIDS validation, start at aperture and then extend to others |
Community projects board
|
Deferred to next meeting due to unclear “Topic” |
Writing up BIDS usage survey into blog post (Franklin) —> Kim |
Some of the analysis code: https://github.com/bids-standard/usage_survey Add info from MRIQC: June 2022: QC metrics from ~375K unique BOLD scans and ~280K T1 scans |
Task definition:
|
Task addition PRs: Anat: https://github.com/bids-standard/bids-specification/pull/1185 - merged PET: https://github.com/bids-standard/bids-specification/pull/1196 - merged Task clarification: https://github.com/bids-standard/bids-specification/issues/1314 - merged |
BIDS derivatives meeting
|
|
Yes or No (maybe just a question for Kim): Can we continue adding Steering Group meeting notes to the BIDS website? |
https://bids.neuroimaging.io/news.html Kim can start doing this again - |
(Added by Eric) Sell BIDS merchandise on TeePublic.com?
|
https://www.teepublic.com/designer-signup https://github.com/bids-standard/bids-specification/tree/master/BIDS_logo https://github.com/bids-standard/bids-specification/blob/master/LICENSE Committee appears to be agreeable with all of this |
Update governance (added by Rémi): Rémi and Eric met to talk about several things.
|
https://bids.neuroimaging.io/governance.html |
Notes
Present: Robert Oostenveld, Guiomar Niso, Eric Earl (visiting maintainer), Yaroslav Halchenko, Kim Ray
Guest: invite Chris Rorden
NOTES from Chris Rorden:
BEPs are done by experts, but implementations are not done by experts
BEPs document examples, but are not clear enough in implementation
details (to entry level users).
Shows unit testing for converting/translating metadata, requires test
datasets and correct outputs
Terminology differs e.g. between Siemens and Philips
-
PET example where dicom standards were not used for BIDS
-
http://www.oneukrainian.com/tmp/DICOMfields.txt has rough grep for descriptions of DICOMs fields matching. Some match some not.
Suggests possibly reaching out to the tool experts/manufacturers to better implement BEPs from start to end.
- Liaisons have been responsive on neurostars (Phillip Clinical Scientists UK Matthew Clemence [github drmclem] Mayo Sandeep Ganji [github sandeepganji])
We have a statement to align BIDS with current standards and vocabularies
- Yarik could not locate it…
SOLUTION:
We cannot assume dcm2nii will fix all of our data conversions.
Reach out to developers for more input on BEPs before they are approved
Provide empty header files for examples of data
- Would be great to have example datasets with sourcedata/ and code/ and README.md provided to describe HOW data was converted/translated
Additional expectations for BEPS before they are approved
- advise each BEP to reach out to manufacturers to make them aware of the standardization ongoing, see the BEP, possibly provide feedback and even may be align metadata/terminology
Consider/formalize a common principle to choose DICOM fields naming in favor of adding new names.
Guio : Lets plan for BIDS Townhall in OpenScience Room - OHBM