We are using BIDS as well as uploading data using XNAT. This causes some confusions, and we’d love to talk with anyone about the general issues that come with integrating the two.
One is this: we code data quality in XNAT (usable vs. unusable, as well as a number from 1 to 4 indicating quality). We don’t want to simply remove this information in the BIDS framework. How do people generally note this? Is there a standard way to note this in the .json file?
Please have a look at this reply to a similar question Poor quality images in BIDS format
Thanks, so much, Chris! I had looked around for comments/discussion about this but hadn’t looked in the right place! I think putting this information in a .tsv file for each subject is a great idea.
This leads to the next question, however, which is, how are people generally dealing with the naming scheme for scans within a sequence of runs (like an AP and a PA), where one of the runs must be re-done because of a quality issue at the time of acquisition. How would one distinguish between two runs collected at the same time without disrupting the over-arching order?
We have toyed with appending “_redo” at the end of the name of the second one, ("_redo2" for the next one, etc.) which isn’t BIDS-compliant. I think this is a fundamental question, and I must be looking in the documentation using the wrong key words!
A concrete example: We have 4 runs of paired AP-PA resting state scans. If we had to re-run the second AP scan, what should we use for the naming convention for that scan? We’d like it to still be associated with the second PA scan by any casual reader of the naming convention.