search

nipreps / nibabies

fMRIPrep-Infants - A robust, transparent workflow tailored for neonate and infant MRI
https://nibabies.readthedocs.io/en/latest
Apache License 2.0
20 stars 10 forks source link

Ensure laplacians are used for registration #35

Open mgxd opened 3 years ago

mgxd commented 3 years ago

They are being fed into antsRegistration https://github.com/nipreps/nibabies/blob/c607811c719cc8e9fed03f9b748cd17818452ddb/nibabies/workflows/anatomical/brain_extraction.py#L274

and show up in the report.rst

* fixed_image : ['/work/nibabies_wf/single_subject_01_wf/infant_anat_wf/infant_brain_extraction_wf/clip_t2w_inu/clipped.nii.gz', '/work/nibabies_wf/single_subject_01_wf/infant_anat_wf/infant_brain_extraction_wf/norm_lap_t2w/clipped.nii.gz']
...
* moving_image : ['/work/nibabies_wf/single_subject_01_wf/infant_anat_wf/infant_brain_extraction_wf/clip_t1w_inu/clipped.nii.gz', '/work/nibabies_wf/single_subject_01_wf/infant_anat_wf/infant_brain_extraction_wf/norm_lap_t1w/clipped.nii.gz']

But are nowhere to be seen in the command:

antsRegistration --collapse-output-transforms 1 --dimensionality 3 --masks /work/nibabies_wf/single_subject_01_wf/infant_anat_wf/infant_brain_extraction_wf/dil_brainmask/tpl-UNCInfant_cohort-1_label-brain_probseg_trans_mask_maths.nii.gz --float 1 --initialize-transforms-per-stage 0 --interpolation LanczosWindowedSinc --output transform --transform Translation[ 1.0 ] --metric Mattes[ /work/nibabies_wf/single_subject_01_wf/infant_anat_wf/infant_brain_extraction_wf/clip_t2w_inu/clipped.nii.gz, /work/nibabies_wf/single_subject_01_wf/infant_anat_wf/infant_brain_extraction_wf/clip_t1w_inu/clipped.nii.gz, 1, 64, Random, 0.5 ] --convergence [ 500, 1e-07, 10 ] --smoothing-sigmas 8.0mm --shrink-factors 2 --use-estimate-learning-rate-once 1 --use-histogram-matching 1 --transform Rigid[ 0.1 ] --metric Mattes[ /work/nibabies_wf/single_subject_01_wf/infant_anat_wf/infant_brain_extraction_wf/clip_t2w_inu/clipped.nii.gz, /work/nibabies_wf/single_subject_01_wf/infant_anat_wf/infant_brain_extraction_wf/clip_t1w_inu/clipped.nii.gz, 1, 64, Random, 0.5 ] --convergence [ 1000x500x250x100, 1e-09, 20 ] --smoothing-sigmas 4.0x2.0x1.0x0.0vox --shrink-factors 8x4x2x1 --use-estimate-learning-rate-once 1 --use-histogram-matching 1 -v --winsorize-image-intensities [ 0.0, 1.0 ] --write-composite-transform 0

mgxd commented 2 years ago

Update:

metric/metric_weight will likely need to be converted to list of lists.

https://github.com/nipreps/nibabies/blob/0c07d6335eba1a136899006659ef18192c316723/nibabies/data/within_subject_t1t2.json#L7-L8