Open pvelasco opened 3 years ago
We could ask Eline how she made these. Or we could discard the eye-tracking files for this session and get a new sample dataset when CBI has tools for incorporating into BIDS.
From Eline: "I made it with a function called edft_to_BIDS.m, it's in the winawerlab git repository MRI_tools > BIDS > edf_to_BIDS.m I wrote it a long time ago, so it might be outdated."
We can either look at that code or just wait for CBI to finish implementing their conversion software. I vote for the latter.
Currently, the
*_physio.json
files are not BIDS compliant.According to the latest stable BIDS spec, the fields
SamplingFrequency
andStartTime
refer to "all columns in the file".Currently, the contents of the json file is a list of dictionaries, with each one of them having a
SamplingFrequency
andStartTime
(which are the same for all elements of the list). Similarly, theFileFormat
andRecordedEye
are the same for all elements.For these files to be BIDS compliant, they should contain a single dictionary with fields
SamplingFrequency
,StartTime
with a single value; and aColumns
field with a list of the column names. E.g., forsub-wlsubj042_ses-01_task-sfp_run-02_physio.json
:You should either correct the original files that are downloaded in
s0
, or add a script ins2
that corrects the*_physio.json
files before running the BIDS validator.Also, note that the
StartTime
is defined as "Start time in seconds in relation to the start of acquisition of the first data sample in the corresponding neural dataset". The current value of 3852115 for the example above would indicate that the eye tracking recording started about 44 days after the first volume of the corresponding functional run...