dcm2bids not ordering runs chronologically about dcm2bids HOT 16 CLOSED

@jcohenadad sorry, I have someone testing this now. Thanks for your patience.

from dcm2bids.

After more investigation, we've noticed that the issue may not be present in SIEMENS but is present for GE and PHILIPS data.

from dcm2bids.

This should be corrected in 2.0.0 but I don't have PHILIPS data to test it

from dcm2bids.

@dasturge can you test it again, I've got some PHILIPS data as well but was not able to replicate your issue.

from dcm2bids.

@arnaudbore I should be able to give it a go in the next week or so, I'll post back with the results.

from dcm2bids.

@dasturge do you have any feedback? Can the issue be closed?

from dcm2bids.

@jcohenadad so unfortunately this doesn't seem to fix our issue.

However, I believe the problem is occurring due to some software mucking up the dicom metadata. I believe the code which does this was acquired directly from PHILIPS, but I'm checking with our collaborators to confirm.

from dcm2bids.

Can you make some dcm from Philips available so we can reproduce and fix it quickly (if possible) ?

from dcm2bids.

@arnaudbore if you sign a Data Use Certification for the ABCD study, we can provide you with a subject's dicoms where this error occurs: https://nda.nih.gov/abcd/request-access

I have confirmed that the GE data where we had the issue was due to a GE script which padded the series number, the PHILIPS data is apparently as-is.

from dcm2bids.

Not sure if someone would like to improve dcm2bids for GE scanner but I can confirm that for GE output, the issue is still there.
To be more precise, SeriesNumber = [3, 4, 11, 6, 5, 10] corresponding to the output of dcm2bids RunNumber = [1, 2, 3, 4, 5, 6]. In this example, SeriesNumber 1 and 2 are omitted, they were loc and anat images. I have not figured out the logic there but it seems the padding issue.
Any suggestions? I'd be happy to provide the data if needed.

from dcm2bids.

@CerenB if you can send me some data I will be happy to fix this issue.

from dcm2bids.

@arnaudbore thank you! it might be due to GE producing extra files and they are messing up the order, still not 100% sure. How can I share the data with you?

from dcm2bids.

@CerenB you can use https://wetransfer.com/.

from dcm2bids.

@CerenB can you test with this PR #68 ? I found something that may be the issue you were talking about.

from dcm2bids.

@CerenB can you test with this PR #68 ? I found something that may be the issue you were talking about.

thanks! it's true that my config.json has upper and lower cases but it's identical to the how Series/Protocol names are called in sidecar. But I am using the latest version of dcm2bids, that should have been resolved in this version?

from dcm2bids.

That PR #68 solved my issue with GE scanner run number organisation. Thank you!

from dcm2bids.