Storage of MEG empty room measurements in the BIDS standard

Community Chat
1

Hi everyone,

My name is Aaron Reer and I am responsible for the data management of the device center for open and reproducible neuroscience tools at the university of Oldenburg. We are currently implementing the BIDS standard for data storage of the two main devices (MRI and MEG scanner) such that all data acquired with the devices will be converted into BIDS conform dataset/directories. Concerning the conversion of MEG data there is something that I stumbled upon which I would like to discuss with the community. It is an issue about the empty-room measurements and their storage.
At our facility many researchers record empty-room measurements for every subject sometimes even before and after the experiment. According to the BIDS specification however, empty-rooms should be stored session-wise in a subject directory sub-emptyroom thus one empty-room recording could be the reference for several subjects. In our case I think it would be more sensible to store the empty-room recording within the meg subdirectory of every subject since there are recordings for every subject.

dataset/sub-001/meg/empty room/pre/
sub-001_task-noisepre_meg.fif
sub-001_task-noisepre_meg.json
sub-001_task-noisepre_channels.tsv

dataset/sub-001/meg/empty room/post/
…

Since I did not work with MEG until I started my position at the university my expertise with respect to MEG is yet very limited and I am also wondering if there is a benefit in measuring the empty room more often and if so if BIDS could be expanded to include directories with such a structure.
I would be glad if somebody could help me on that matter.

Cheers,
Aaron

1 Like
2

@robert has probably given this some thought, maybe he can chime in.

3

This way to organize empty rooms recording is RECOMMENDED but not REQUIRED.

So you could organize things differently and still comply with BIDS.

└── sub-01
    ├── ses-emptyRoomPost    # camelCase here to make it more human readable
    ├── ses-emptyRoomPre
    └── ses-experiment       # can be called something else than experiment

Remember that session folders can be used for “logical” grouping and that things that happened on the same day can totally be put in different folders: you can then keep track of what happened when by logging this info in the sessions.tsv and scans.tsv