DOC: ME | Add missing output to documentation, improve boilerplate #2608
Conversation
Codecov Report
@@ Coverage Diff @@
## master #2608 +/- ##
=======================================
Coverage 45.80% 45.80%
=======================================
Files 41 41
Lines 3133 3133
=======================================
Hits 1435 1435
Misses 1698 1698
Continue to review full report at Codecov.
|
fmriprep/workflows/bold/t2s.py
Outdated
| Combine multiple echos of :abbr:`ME-EPI (multi-echo echo-planar imaging)`. | ||
|
|
||
| This workflow wraps the `tedana`_ `T2* workflow`_ to optimally | ||
| combine multiple echos and derive a T2* map. | ||
| combine multiple echos and derive a T2\ :sup:`★` map. | ||
| The following steps are performed: | ||
|
|
||
| #. :abbr:`HMC (head motion correction)` on individual echo files. |
There was a problem hiding this comment.
Now that we are at it. I've been searching through tedana's docs and didn't see that the workflow runs any sort of HMC internally, is that correct?
If it does, I would need to confirm we are not running HMC twice in fMRIPrep. In that case, would it be possible to disable HMC in tedana?
If it doesn't, I think I will revise this docstring to say that HMC'd echos are assumed (i.e., the workflow does not run HMC again).
There was a problem hiding this comment.
tedana doesn't perform any HMC, so I agree that the docstring needs to be revised.
Although I brought up the issue within #2530 and got some mildly opposed initial opinion from @effigies (#2530 (comment)), I've come to think that we should provide the T2star workflow with HMC'ed and SDC'ed echos (confirmation awaiting: #2608 (review)). This comes as an issue related to #2606. This PR proposes that individual echoes are HMC'ed (and SDC'ed when fieldmaps are available) before calculating the optimal combination and the T2star map. I believe this may be beneficial when we combine SDC with modulation by the Jacobian of the distortion (nipreps/sdcflows#238) because that might help restore some (little, admittedly) of the dropout of the later echoes in distorted regions. I would expect more voxels will be deemed as acceptable for the exponential fitting for this reason. It also conceptually simplifies a little, as both SE and ME paths look more alike, and only the little hacks to generate ME iterables and the ``boldbuffer`` are now necessary. Otherwise, SDC would be inserted at a much later stage for ME only.
Co-authored-by: Taylor Salo <tsalo006@fiu.edu>
Document one missing output of the T2star workflow, and edit boilerplate.