Storing data on participant groups in bids-compliant file structure

Dear authors,
In our project amongst other topics we work on ensuring data storage compliant with the BIDS standard. While mne-bids allows to store such information as task, session, age etc, we could not find a clear way of storing the “group” marker (e.g. left-handed, right-handed etc). Could you suggest us the most meaningful way of storing such kind of information? A code example containing some mne-bids tools would be most appreciated :smiley:
Thanks in advance
Best regards,
Michał

hi,

This is typically written in the participants.tsv file.

You can put this information in the raw.info[“subject_info”][“hand”]. See
https://mne.tools/dev/generated/mne.Info.html#mne.Info

Alex

1 Like

@agramfort’s answer is the correct way to do it.

Actually I believe we currently don’t cover this in any examples of the MNE-BIDS docs. But we should :sweat_smile:

@sappelhoff thoughts?

Thank you for your ansewer. Maybe the example was not very good. By a “group” I meant not only handedness, but any arbitrary group (the examples may be more or less abstract: student of IT/ student of maths/student of psychology, education: only school education/university bachelor/university master/university PhD etc).

These also go into participants.tsv as custom columns and, ideally, are accompanied by an explanation of their meaning in participants.json. See the BIDS spec:
https://bids-specification.readthedocs.io/en/stable/03-modality-agnostic-files.html#participants-file

I don’t believe MNE-BIDS currently allows adding custom columns via its API, but it will respect (and preserve) anything you’ve added manually.

Richard

1 Like

indeed, you’ll have to write some custom code for this, e.g., using pandas to read in the participants file and then editing it.

I’d do that as a last step, I think.

1 Like

I believe there is an open issue for this: https://github.com/mne-tools/mne-bids/issues/439. This may be worth revisiting …

Mainak

1 Like

Thank you for your responses! I think we will try to implement saving such kind of data into participants.tsv.