Open sitek opened 1 year ago
@sitek and @bpinsard - do you have samples of such peculiar data to share? (ideally openly, may be some phantoms)
Hi @sitek & @yarikoptic,
I'm running into a comparable problem and was wondering if there are any updates concerning this?
Cheers, Peer
Summary
Using a CMRR Spin Echo EPI sequence for PEPolar SDC correction on a 7T. We're collecting SE EPI before every BOLD run.
SBRef
is exported separately:So I exclude that
series_description
text in myheuristic.py
:And—correctly—the SE EPI
SBRef
doesn't appear in*auto.txt
– here's the output ofcat .heudiconv/FLT10/ses-1/info/FLT10_ses-1.auto.txt
:But heudiconv converts the
SBRef
files anyway and gives thempart-{mag|phase}
descriptors:with
sub-FLT10_acq-func_dir-AP_run-01_part-mag_epi.json
noting that they come fromSBRef files
:"SeriesDescription": "FieldMap_SE_AP_Tone_Learning_1_SBRef",
The files aren't ordered the same between AP and PA acquisitions, either—for instance,
dir-AP_run-03
is a true SE EPI magnitude image, whiledir-PA_run-03
is anSBRef
part-{mag|phase}
pair. This then throws offIntendedFor
matching, since there's nodir-PA
equivalent todir-AP_run-03
, so one of mybold
runs doesn't get included in anydir-PA
IntendedFor
fields.May be related to #549 and associated PRs.
Platform details:
Choose one:
[ X] Local environment
[ ] Container
Heudiconv version: 0.11.6, but existed in 0.9.x versions as well (which is all that I've tested)