registration_synthmorph

Perform registration using SynthMorph from Freesurfer. Outputs transforms in Freesurfer format .lta for affine and .nii.gz (synthmorph also supports .mgz) for deform, both in RAS orientation. Conversion to other formats is done using lta_convert and mri_warp_convert respectively, which support a wide range of conversion formats and orientations, such as ANTs ans FSL. To convert the output of this module, use the registration/convert module successively to it. Note that tests using synthmorph are non-reproductible.

Keywords : Registration, Brain imaging, MRI, Synthetic, AI, CNN

---

Inputs

Input 1

Format : tuple val(meta), path(fixed_image), path(moving_image)

	Type	Description	Mandatory	Pattern
meta	map	Groovy Map containing sample information e.g. [ id:'test', single_end:false ]	True
fixed_image	file	Nifti volume fixed for registration	True	*.{nii,nii.gz}
moving_image	file	Nifti volume moving for registration	True	*.{nii,nii.gz}

---

Outputs

image_warped

Format : tuple val(meta), path(*_warped.nii.gz)

	Type	Description	Mandatory	Pattern
meta	map	Groovy Map containing sample information e.g. [ id:'test', single_end:false ]	True
*_warped.nii.gz	file	Warped image	True	*_warped.nii.gz

forward_affine

Format : tuple val(meta), path(*_forward{0,1,_standalone}_affine.lta)

	Type	Description	Mandatory	Pattern
meta	map	Groovy Map containing sample information e.g. [ id:'test', single_end:false ]	True
*_forward{0,1,_standalone}_affine.lta	file	Affine transformation matrix to fixed space.	True	*_forward{0,1,_standalone}_affine.lta

forward_warp

Format : tuple val(meta), path(*_forward0_deform.nii.gz)

	Type	Description	Mandatory	Pattern
meta	map	Groovy Map containing sample information e.g. [ id:'test', single_end:false ]	True
*_forward0_deform.nii.gz	file	Deformation field to fixed space.	True	*_forward0_deform.nii.gz

backward_warp

Format : tuple val(meta), path(*_backward1_deform.nii.gz)

	Type	Description	Mandatory	Pattern
meta	map	Groovy Map containing sample information e.g. [ id:'test', single_end:false ]	True
*_backward1_deform.nii.gz	file	Deformation field to moving space.	True	*_backward1_deform.nii.gz

backward_affine

Format : tuple val(meta), path(*_backward{0,_standalone}_affine.lta)

	Type	Description	Mandatory	Pattern
meta	map	Groovy Map containing sample information e.g. [ id:'test', single_end:false ]	True
*_backward{0,_standalone}_affine.lta	file	Affine transformation matrix to moving space.	True	*_backward{0,_standalone}_affine.lta

forward_image_transform

Format : tuple val(meta), val(*_forward[!_]*.{lta,nii.gz})

	Type	Description	Mandatory	Pattern
meta	map	Groovy Map containing sample information e.g. [ id:'test', single_end:false ]	True
forward[!].{lta,nii.gz}	list	Tuple, transformation files to warp images to fixed space, in the correct order for REGISTRATION_ANTSAPPLYTRANSFORMS : [ meta, [ forward_warp, forward_affine ] ].	True	*_forward[!_]*.{lta,nii.gz}

backward_image_transform

Format : tuple val(meta), val(*_backward[!_]*.{lta,nii.gz})

	Type	Description	Mandatory	Pattern
meta	map	Groovy Map containing sample information e.g. [ id:'test', single_end:false ]	True
backward[!].{lta,nii.gz}	list	Tuple, transformation files to warp images to moving space, in the correct order for REGISTRATION_ANTSAPPLYTRANSFORMS : [ meta, [ backward_affine, backward_warp ] ].	True	*_backward[!_]*.{lta,nii.gz}

forward_tractogram_transform

Format : tuple val(meta), val(*_backward[!_]*.{lta,nii.gz})

	Type	Description	Mandatory	Pattern
meta	map	Groovy Map containing sample information e.g. [ id:'test', single_end:false ]	True
backward[!].{lta,nii.gz}	list	Tuple, transformation files to warp tractograms to fixed space, in the correct order for REGISTRATION_TRANSFORMTRACTOGRAM : [ meta, [ backward_affine, backward_warp ] ].	True	*_backward[!_]*.{lta,nii.gz}

backward_tractogram_transform

Format : tuple val(meta), val(*_forward[!_]*.{lta,nii.gz})

	Type	Description	Mandatory	Pattern
meta	map	Groovy Map containing sample information e.g. [ id:'test', single_end:false ]	True
forward[!].{lta,nii.gz}	list	Tuple, transformation files to warp tractograms to moving space, in the correct order for REGISTRATION_TRANSFORMTRACTOGRAM : [ meta, [ forward_warp, forward_affine ] ].	True	*_forward[!_]*.{lta,nii.gz}

versions

Format : path(versions.yml)

	Type	Description	Mandatory	Pattern
versions.yml	file	File containing software versions	True	versions.yml

---

Arguments (see process.ext)

	Type	Description	Default	Choices
models	list	List of transformation models to apply to the moving image in series.	[“affine”, “deform”]	- rigid - affine - deform - joint
weights	list	List of weights to apply to each model. Must supply one per model on none. Use “none” to specify no weight for a given model. When using the “joint” model, you must add 2 weights in the list for affine and warp.
use_gpu	boolean	Use GPU to accelerate computations (must enable GPU support).	False
regularization	float	Regularizationto add to deformation steps.	0.5
steps	integer	Number of iteration for the optimisation of deformations.	7
extent	integer	Size of the deformation grid in voxels.	256	- 192 - 256
disable_resampling	boolean	Update image header instead of performing resampling of the image grid.	False

---

Tools

	Description	DOI
Freesurfer	Open source neuroimaging toolkit for processing, analyzing, and visualizing human brain MR images.	10.1016/j.neuroimage.2012.01.021
SynthMorph	Synthmorph registration method	10.1109/TMI.2021.3116879

---

Authors

• @anroy1
• @AlexVCaron

---

Last updated : 2025-11-03