BIDS validator error for dwi

1

Hello

I used the online BIDs validator on my dataset, and I have a problem where few of the subjects have dwi files that are shorter than others.
To accomodate this, I have a ‘dwi.bval’, and ‘dwi.bvec’ file at the BIDS root folder which applies to all dwi files, and added additional bval files and bvec files just for the subjects who have shorter dwi files (e.g., BIDS\sub-dmp0110\ses-scan2\dwi\sub-dmp0110_ses-scan2_dwi.bval) . I was under the assumption that this should work on the inheritance principle.
The online BIDS validator gives me errors for volume count mismatch:
The number of volumes in this scan does not match the number of volumes in the corresponding .bvec and .bval files.

When I’m running fmriprep, this issue doesn’t come up so I’m assuming that the BIDS validator included in fmriprep doesn’t check for dwi files…

Is the best procedure to include separate bval and bvec files for every subject separately?

2

Hi @sangillee3rd

Thank you for your message! By chance were the files that were flagged by the error all matched those that had a bval and bvec defined at the modality level? From what I understand this appears to be a valid implementation of the inheritance principle

Thank you,
Franklin

cc: @rwblair

3

Hi @franklin, Yes that’s the case.
To be more specific, I initially didn’t know that some subjects had dwis with fewer volumes. So I only included the bval and bvec files at the BIDS root directory and the BIDS validator (online) thankfully caught that few dwi files were short on volume. However, even after I supplied the flagged dwi files with their own bval bvec (and also tried adding their own json files), they were still being flagged for the same error.

Thank you
Arthur

4

Hi @sangillee3rd

Thank you for your message! By chance were these the only errors you saw? Errors could be associated with each other and causing each other.

Perhaps another approach (or sanity check) can be to copy your top level bval bvec into the subject dwi folders with the respective file names. This would be teasing apart the inheritance principle bug from a validated BIDS dataset.

Thank you,
Franklin

5

@franklin
Thanks for the quick reply!
Yes they were the only errors that I saw.
I also did try specifying a separate bval and bvec file for each subject instead of using the root directory files and that removed the errors. It just has a lot of redundant files.

Arthur

6

Hi @sangillee3rd

Sorry for the delayed response! Good to hear the work around worked - we are working on patching the inheritance bug

Thank you,
Franklin