ENH: Add qMRI example datasets

[effigies]

Adapted examples from https://osf.io/k4bs5/ for BEP001.

For easier review, here are the dataset trees:

• https://github.com/effigies/BIDS-examples/tree/qmri/qmri_mp2rage
• https://github.com/effigies/BIDS-examples/tree/qmri/qmri_mp2rageme
• https://github.com/effigies/BIDS-examples/tree/qmri/qmri_mpm
• https://github.com/effigies/BIDS-examples/tree/qmri/qmri_mtsat
• https://github.com/effigies/BIDS-examples/tree/qmri/qmri_qsm
• https://github.com/effigies/BIDS-examples/tree/qmri/qmri_sa2rage
• https://github.com/effigies/BIDS-examples/tree/qmri/qmri_vfa

[effigies]

Getting errors about fmap/*_TB1map.nii. Should these be moved to anat/ or should the spec be updated to put them in fmap/?

cc @agahkarakuzu @tsalo (please cc anybody else who should be added)

[tsalo]

They should be under fmap (see here), but I missed them in https://github.com/bids-standard/bids-validator/pull/1161. 🤦

I can open a PR to the validator to fix that oversight.

EDIT: I've opened bids-standard/bids-validator#1183.

[effigies]

@tsalo To be clear, I meant TB1Map/RB1Map (schema). Those are declared in anatomical.

[effigies]

Nice catches, @sappelhoff. I think we're down to

• Pending final decisiions for [SCHEMA] Update qMRI fieldmap schema bids-specification#728 and a corresponding validator PR.
• Clarification on whether T/RB1Maps should be fieldmap or anat and how to encode that in the spec, and a corresponding validator PR.

[sappelhoff]

We are getting a warning for the TotalReadoutTime fields in the qmri_mpm datasets:

Seems valid:

[sappelhoff]

we also get warnings about missing Author fields. I inquired with the BEP001 team how to resolve this: https://mattermost.brainhack.org/brainhack/pl/icwy7hqqwtnjzjrzpmmgmehgby

Probably adding something like ["qmri BEP Team", "BIDS maintainers"] will be fine.

[Gilles86]

@sappelhoff ,

The MP2RAGE and MP2RAGE-ME datasets are authored by myself and Tomas Knapen.

For the MPM I think you should check with @ChristophePhillips and @lazaral

I think the rest has been provided by @agahkarakuzu

[agahkarakuzu]

@sappelhoff I noticed that the validator was complaining about single authorship :) I started expanding the example datasets and validator, I can edit them as I go by preserving 1st dataset providers' order.

[sappelhoff]

Thanks both, I think it'd be best if you could update the OSF datasets with the authors, and mirror the changes that @effigies already did (see e.g., 7e1ec55 and 8d6f42d) --> Then Chris can pull these changes into his datalad controlled datasets, and then push them to bids-examples.

Perhaps you can also solve #245 (comment) as you go

as for validator, please see: https://github.com/bids-standard/bids-validator/pull/1186

from that PR, the validator now passes for qMRI.

[lazaral]

@sappelhoff, regarding the MPM dataset, we used the one from this publication: https://www.sciencedirect.com/science/article/pii/S235234091930486X?via%3Dihub

We currently have the first author of that manuscript names as author of the dataset in the OSF repo. Should we add the full author list to the dataset_description.json?

[sappelhoff]

Should we add the full author list to the dataset_description.json?

@lazaral yes please, make it an array of strings: ["author1", "author2", ...]

[sappelhoff]

All datasets are passing bids-validator when validating using the changes from https://github.com/bids-standard/bids-validator/pull/1186

As I mention in https://github.com/bids-standard/bids-validator/pull/1186#issuecomment-779333369, I suggest to merge this PR into the https://github.com/bids-standard/bids-examples/tree/bep001_qmri branch.

Then we can sort out the final issues and them merge that one into master