We acquired data from the same ieeg patient with two separate amplifiers: 1 amplifier recorded from a 1-cm spaced grid and 1 amplifier recorded from a 3-mm spaced grid. The start time, sampling frequency and filter settings are different for the two amplifiers but both amplifiers recorded the same task in the same patient. The event file would be the same, because the patient did the same task.
How can I keep the two files separate in BIDS/ieeg? Looking at the doc the only naming structure allowed is sub-<label>[_ses-<label>]_task-<label>[_run-<index>]_ieeg.eeg. I see there is an option to have simultaneous recordings but I’m not sure how to differentiate it here.
Another option is to keep them separate using an additional field, such as acq or recording
Do you have any suggestions on how to store data from two different amplifiers?
Thank you
In addition, I would make heavy use of the README file in this case.
I think it’d be cool to also document this case in the bids-starter-kit as a tutorial, @gpiantoni perhaps you’d be willing to add a short section, describing your case and solution.
When drafting the electrophys extensions we did explicitly consider EEG and MEG using different systems (which would go in separate directories), but not two times the same modality. But I think there is room for it: if you look at https://bids-specification.readthedocs.io/en/stable/99-appendices/04-entity-table.html you can see that acq is optional.
I suggest sub-01_task-rest_acq-highres_ieeg.edf sub-01_task-rest_acq-highres_ieeg.json sub-01_task-rest_acq-highres_events.tsv sub-01_task-rest_acq-highres_channels.tsv sub-01_task-rest_acq-highres_electrodes.tsv
and sub-01_task-rest_acq-lowres_ieeg.edf sub-01_task-rest_acq-lowres_ieeg.json sub-01_task-rest_acq-lowres_events.tsv sub-01_task-rest_acq-lowres_channels.tsv sub-01_task-rest_acq-lowres_electrodes.tsv
A bit annoying might be that the electrode positions (which you are likely to get from a CT) would have to be stored in two separate files. But other than that I think this is a logical naming scheme. It would also allow cases where you only have one of the two (e.g. when the other amplifier was not running) or when the start times of the amplifiers are very out of sync.
In the scans.tsv you would have to indicate the acquisition start time, so that the two can be read and aligned. And/or you might have events that are shared between amplifiers: those would be in both events files.
Thank you for the pointers. I should have checked the Entity Table, as Robert suggested, instead of looking only at the ieeg-specific page. I created a PR to make the modality-specific pages consistent with the Entity Table.
Note that according to the Entity Table, acq-<acquisition> is not an optional field for events.tsv, though