ENH: Connect fieldmap estimation to preprocessing pipeline of individual DWI runs #140
Conversation
|
Hello @josephmje, Thank you for updating! Cheers! There are no style issues detected in this Pull Request. 🍻 To test for issues locally, Comment last updated at 2020-12-11 14:50:57 UTC |
josephmje
force-pushed
the
enh/apply_topup
branch
from
d0c7912
to
ab515da
Compare
josephmje
force-pushed
the
enh/apply_topup
branch
from
ab515da
to
2edc4b3
Compare
oesteban
changed the title
Allows for an interface similar to that of *sMRIPrep*, which enables using ``KeySelect`` nodes to find the appropriate connections downstream. Required-by: nipreps/dmriprep#140. Resolves: nipreps#147.
oesteban
force-pushed
the
enh/apply_topup
branch
from
3f12721
to
b750bbb
Compare
- Connections to the particular ``dwi_file`` are done within the particular run's preproc workflow, using ``KeySelect`` to demux the available fieldmap options. This relies on implementing nipreps/sdcflows#147 and nipreps/sdcflows#148. - Minimizes the overhead in ``dmriprep/workflows/base.py`` - Run black
oesteban
force-pushed
the
enh/apply_topup
branch
from
346dddd
to
30039bb
Compare
|
@oesteban Thanks for the refactor to the base workflow. Reading it now and it looks a lot more intuitive. |
josephmje
force-pushed
the
enh/apply_topup
branch
from
9711a19
to
9a6b7d5
Compare
…EPI target This code adds a ``get_identifier()`` interface that allows the user find the a list of ``B0FieldIdentifier``(s) that are associated with a given file. ATM, it pulls the information from the ``IntendedFor`` metadata of fieldmaps, but in the future, it should first read the ``B0FieldSource``. Originated-by: nipreps/dmriprep#140. Resolves: nipreps#148.
|
@josephmje can you make sure you have pushed anything you were working on and hold off of pushing while I finish the SDCFlows updates? I'll update this branch when SDCFlows is ready. |
Yup everything I've been working on has been pushed. |
oesteban
force-pushed
the
enh/apply_topup
branch
from
ebeb2fd
to
9371c49
Compare
|
Okay, if tests pass then I think we can go ahead and merge. It still won't do anything with the fieldmaps because our test dataset with fieldmaps (ds001771) does not have But I think that's a different issue (probably pertaining to SDCFlows, and namely the problem of what to do when no IntendedFor are given for a potential EPI target). And, as currently implemented, the behavior is similar to fMRIPrep's (which does nothing if IntendedFor is not set). |
|
Should we update the docstring for these workflows so that the fieldmap workflows get visualized? Currently, it refers to |
Good luck with that :p It's going to be a rabbit hole of massive dimensions. Plus, SDCFlows has ~94% coverage. Let's trust that. |
No description provided.