def ants_registration_affine_node(**kwargs):
"""return antsRegistration interace instance with default values
based on antsRegistrationSyN.sh with the a transformation option
:param \\*\\*kwargs: parameters to override the default values
:return: :py:obj:`Registration` node
"""
defaults = dict(dimension=3,
use_histogram_matching=False,
interpolation='Linear',
initial_moving_transform_com=1,
metric=['MI', 'MI'],
metric_weight=[1.0, 1.0],
radius_or_number_of_bins=[32, 32],
sampling_strategy=['Regular', 'Regular'],
sampling_percentage=[0.25, 0.25],
transforms=['Rigid', 'Affine'],
transform_parameters=[(0.1, ), (0.1, )],
smoothing_sigmas=[[3, 2, 1, 0], [3, 2, 1, 0]],
sigma_units=['vox', 'vox'],
shrink_factors=[[8, 4, 2, 1], [8, 4, 2, 1]],
number_of_iterations=[[1000, 500, 250, 100],
[1000, 500, 250, 100]],
convergence_threshold=[1e-6, 1e-6],
convergence_window_size=[10, 10],
winsorize_lower_quantile=0.005,
winsorize_upper_quantile=0.995,
write_composite_transform=True,
output_warped_image=True)
defaults.update(kwargs)
ants = Registration(**defaults)
return ants
def ants_registration_syn_no_affine_node(**kwargs):
"""return antsRegistration interace instance with default values
based on antsRegistrationSyN.sh with the s transformation option
and the rigid and affine steps removed
:param \\*\\*kwargs: parameters to override the default values
:return: :py:obj:`Registration` node
"""
defaults = dict(dimension=3,
use_histogram_matching=False,
interpolation='Linear',
metric=['CC'],
metric_weight=[1.0],
radius_or_number_of_bins=[4],
sampling_strategy=[None],
sampling_percentage=[None],
transforms=['SyN'],
transform_parameters=[(0.1, 3, 0)],
smoothing_sigmas=[[3, 2, 1, 0]],
sigma_units=['vox'],
shrink_factors=[[8, 4, 2, 1]],
number_of_iterations=[[100, 70, 50, 20]],
convergence_threshold=[1e-6],
convergence_window_size=[10],
winsorize_lower_quantile=0.005,
winsorize_upper_quantile=0.995,
write_composite_transform=True,
output_warped_image=True)
defaults.update(kwargs)
ants = Registration(**defaults)
return ants
def _config_ants(self, ants_settings):
NIWORKFLOWS_LOG.info('Loading settings from file %s.', ants_settings)
self.norm = Registration(
moving_image=self.inputs.moving_image,
num_threads=self.inputs.num_threads,
from_file=ants_settings,
terminal_output='file',
write_composite_transform=True
)
if isdefined(self.inputs.moving_mask):
if self.inputs.explicit_masking:
self.norm.inputs.moving_image = mask(
self.inputs.moving_image[0],
self.inputs.moving_mask,
"moving_masked.nii.gz")
else:
self.norm.inputs.moving_image_mask = self.inputs.moving_mask
if isdefined(self.inputs.reference_image):
self.norm.inputs.fixed_image = self.inputs.reference_image
if isdefined(self.inputs.reference_mask):
if self.inputs.explicit_masking:
self.norm.inputs.fixed_image = mask(
self.inputs.reference_image[0],
self.inputs.mreference_mask,
"fixed_masked.nii.gz")
else:
self.norm.inputs.fixed_image_mask = self.inputs.reference_mask
else:
get_template = getattr(getters, 'get_{}'.format(self.inputs.template))
mni_template = get_template()
if self.inputs.orientation == 'LAS':
raise NotImplementedError
resolution = self.inputs.template_resolution
if self.inputs.testing:
resolution = 2
if self.inputs.explicit_masking:
self.norm.inputs.fixed_image = mask(op.join(
mni_template, '%dmm_%s.nii.gz' % (resolution, self.inputs.reference)),
op.join(
mni_template, '%dmm_brainmask.nii.gz' % resolution),
"fixed_masked.nii.gz")
else:
self.norm.inputs.fixed_image = op.join(
mni_template,
'%dmm_%s.nii.gz' % (resolution, self.inputs.reference))
self.norm.inputs.fixed_image_mask = op.join(
mni_template, '%dmm_brainmask.nii.gz' % resolution)
def test_Registration_outputs():
output_map = dict(composite_transform=dict(),
forward_invert_flags=dict(),
forward_transforms=dict(),
inverse_composite_transform=dict(),
inverse_warped_image=dict(),
reverse_invert_flags=dict(),
reverse_transforms=dict(),
warped_image=dict(),
)
outputs = Registration.output_spec()
for key, metadata in output_map.items():
for metakey, value in metadata.items():
yield assert_equal, getattr(outputs.traits()[key], metakey), value
def test_Registration_outputs():
output_map = dict(
composite_transform=dict(),
forward_invert_flags=dict(),
forward_transforms=dict(),
inverse_composite_transform=dict(),
inverse_warped_image=dict(),
reverse_invert_flags=dict(),
reverse_transforms=dict(),
warped_image=dict(),
)
outputs = Registration.output_spec()
for key, metadata in output_map.items():
for metakey, value in metadata.items():
yield assert_equal, getattr(outputs.traits()[key], metakey), value
def test_Registration_inputs():
input_map = dict(args=dict(argstr='%s',
),
collapse_linear_transforms_to_fixed_image_header=dict(argstr='%s',
usedefault=True,
),
collapse_output_transforms=dict(argstr='--collapse-output-transforms %d',
usedefault=True,
),
convergence_threshold=dict(requires=['number_of_iterations'],
usedefault=True,
),
convergence_window_size=dict(requires=['convergence_threshold'],
usedefault=True,
),
dimension=dict(argstr='--dimensionality %d',
usedefault=True,
),
environ=dict(nohash=True,
usedefault=True,
),
fixed_image=dict(mandatory=True,
),
fixed_image_mask=dict(argstr='%s',
),
ignore_exception=dict(nohash=True,
usedefault=True,
),
initial_moving_transform=dict(argstr='%s',
xor=['initial_moving_transform_com'],
),
initial_moving_transform_com=dict(argstr='%s',
xor=['initial_moving_transform'],
),
interpolation=dict(argstr='%s',
usedefault=True,
),
invert_initial_moving_transform=dict(requires=['initial_moving_transform'],
xor=['initial_moving_transform_com'],
),
metric=dict(mandatory=True,
),
metric_item_trait=dict(),
metric_stage_trait=dict(),
metric_weight=dict(mandatory=True,
requires=['metric'],
usedefault=True,
),
metric_weight_item_trait=dict(),
metric_weight_stage_trait=dict(),
moving_image=dict(mandatory=True,
),
moving_image_mask=dict(requires=['fixed_image_mask'],
),
num_threads=dict(nohash=True,
usedefault=True,
),
number_of_iterations=dict(),
output_inverse_warped_image=dict(hash_files=False,
requires=['output_warped_image'],
),
output_transform_prefix=dict(argstr='%s',
usedefault=True,
),
output_warped_image=dict(hash_files=False,
),
radius_bins_item_trait=dict(),
radius_bins_stage_trait=dict(),
radius_or_number_of_bins=dict(requires=['metric_weight'],
usedefault=True,
),
sampling_percentage=dict(requires=['sampling_strategy'],
),
sampling_percentage_item_trait=dict(),
sampling_percentage_stage_trait=dict(),
sampling_strategy=dict(requires=['metric_weight'],
),
sampling_strategy_item_trait=dict(),
sampling_strategy_stage_trait=dict(),
shrink_factors=dict(mandatory=True,
),
sigma_units=dict(requires=['smoothing_sigmas'],
),
smoothing_sigmas=dict(mandatory=True,
),
terminal_output=dict(mandatory=True,
nohash=True,
),
transform_parameters=dict(),
transforms=dict(argstr='%s',
mandatory=True,
),
use_estimate_learning_rate_once=dict(),
use_histogram_matching=dict(usedefault=True,
),
winsorize_lower_quantile=dict(argstr='%s',
usedefault=True,
),
winsorize_upper_quantile=dict(argstr='%s',
usedefault=True,
),
write_composite_transform=dict(argstr='--write-composite-transform %d',
usedefault=True,
),
)
inputs = Registration.input_spec()
for key, metadata in input_map.items():
for metakey, value in metadata.items():
yield assert_equal, getattr(inputs.traits()[key], metakey), value
def fit(
dwdata,
*,
n_iter=1,
align_kwargs=None,
model="b0",
omp_nthreads=None,
seed=None,
**kwargs,
):
r"""
Estimate head-motion and Eddy currents.
Parameters
----------
dwdata : :obj:`~eddymotion.dmri.DWI`
The target DWI dataset, represented by this tool's internal
type. The object is used in-place, and will contain the estimated
parameters in its ``em_affines`` property, as well as the rotated
*b*-vectors within its ``gradients`` property.
n_iter : :obj:`int`
Number of iterations this particular model is going to be repeated.
align_kwargs : :obj:`dict`
Parameters to configure the image registration process.
model : :obj:`str`
Selects the diffusion model that will generate the registration target
corresponding to each gradient map.
See :obj:`~eddymotion.model.ModelFactory` for allowed models (and corresponding
keywords).
seed : :obj:`int` or :obj:`bool`
Seed the random number generator (necessary when we want deterministic
estimation).
Return
------
affines : :obj:`list` of :obj:`numpy.ndarray`
A list of :math:`4 \times 4` affine matrices encoding the estimated
parameters of the deformations caused by head-motion and eddy-currents.
"""
align_kwargs = align_kwargs or {}
reg_target_type = ("dwi" if model.lower() not in ("b0", "s0", "avg",
"average",
"mean") else "b0")
if seed or seed == 0:
np.random.seed(20210324 if seed is True else seed)
bmask_img = None
if dwdata.brainmask is not None:
_, bmask_img = mkstemp(suffix="_bmask.nii.gz")
nb.Nifti1Image(dwdata.brainmask.astype("uint8"), dwdata.affine,
None).to_filename(bmask_img)
kwargs["mask"] = dwdata.brainmask
kwargs["S0"] = _advanced_clip(dwdata.bzero)
if "n_threads" in kwargs:
align_kwargs["num_threads"] = kwargs["n_threads"]
for i_iter in range(1, n_iter + 1):
index_order = np.arange(len(dwdata))
np.random.shuffle(index_order)
with tqdm(total=len(index_order), unit="dwi") as pbar:
for i in index_order:
# run a original-to-synthetic affine registration
with TemporaryDirectory() as tmpdir:
pbar.write(
f"Pass {i_iter}/{n_iter} | Processing b-index <{i}> in <{tmpdir}>"
)
data_train, data_test = dwdata.logo_split(i,
with_b0=True)
# Factory creates the appropriate model and pipes arguments
dwmodel = ModelFactory.init(gtab=data_train[1],
model=model,
omp_nthreads=omp_nthreads,
**kwargs)
# fit the model
dwmodel.fit(data_train[0])
# generate a synthetic dw volume for the test gradient
predicted = dwmodel.predict(data_test[1])
tmpdir = Path(tmpdir)
moving = tmpdir / "moving.nii.gz"
fixed = tmpdir / "fixed.nii.gz"
_to_nifti(data_test[0], dwdata.affine, moving)
_to_nifti(
predicted,
dwdata.affine,
fixed,
clip=reg_target_type == "dwi",
)
registration = Registration(
terminal_output="file",
from_file=pkg_fn(
"eddymotion",
f"config/dwi-to-{reg_target_type}_level{i_iter}.json",
),
fixed_image=str(fixed.absolute()),
moving_image=str(moving.absolute()),
**align_kwargs,
)
if bmask_img:
registration.inputs.fixed_image_masks = [
"NULL", bmask_img
]
if dwdata.em_affines and dwdata.em_affines[
i] is not None:
mat_file = tmpdir / f"init{i_iter}.mat"
dwdata.em_affines[i].to_filename(mat_file,
fmt="itk")
registration.inputs.initial_moving_transform = str(
mat_file)
# execute ants command line
result = registration.run(cwd=str(tmpdir)).outputs
# read output transform
xform = nt.io.itk.ITKLinearTransform.from_filename(
result.forward_transforms[0]).to_ras(
reference=fixed, moving=moving)
# update
dwdata.set_transform(i, xform)
pbar.update()
return dwdata.em_affines
def test_Registration_inputs():
input_map = dict(
args=dict(argstr='%s', ),
collapse_output_transforms=dict(
argstr='--collapse-output-transforms %d',
usedefault=True,
),
convergence_threshold=dict(
requires=['number_of_iterations'],
usedefault=True,
),
convergence_window_size=dict(
requires=['convergence_threshold'],
usedefault=True,
),
dimension=dict(
argstr='--dimensionality %d',
usedefault=True,
),
environ=dict(
nohash=True,
usedefault=True,
),
fixed_image=dict(mandatory=True, ),
fixed_image_mask=dict(argstr='%s', ),
float=dict(argstr='--float %d', ),
ignore_exception=dict(
nohash=True,
usedefault=True,
),
initial_moving_transform=dict(
argstr='%s',
xor=['initial_moving_transform_com'],
),
initial_moving_transform_com=dict(
argstr='%s',
xor=['initial_moving_transform'],
),
initialize_transforms_per_stage=dict(
argstr='--initialize-transforms-per-stage %d',
usedefault=True,
),
interpolation=dict(
argstr='%s',
usedefault=True,
),
invert_initial_moving_transform=dict(
requires=['initial_moving_transform'],
xor=['initial_moving_transform_com'],
),
metric=dict(mandatory=True, ),
metric_item_trait=dict(),
metric_stage_trait=dict(),
metric_weight=dict(
mandatory=True,
requires=['metric'],
usedefault=True,
),
metric_weight_item_trait=dict(),
metric_weight_stage_trait=dict(),
moving_image=dict(mandatory=True, ),
moving_image_mask=dict(requires=['fixed_image_mask'], ),
num_threads=dict(
nohash=True,
usedefault=True,
),
number_of_iterations=dict(),
output_inverse_warped_image=dict(
hash_files=False,
requires=['output_warped_image'],
),
output_transform_prefix=dict(
argstr='%s',
usedefault=True,
),
output_warped_image=dict(hash_files=False, ),
radius_bins_item_trait=dict(),
radius_bins_stage_trait=dict(),
radius_or_number_of_bins=dict(
requires=['metric_weight'],
usedefault=True,
),
restore_state=dict(argstr='--restore-state %s', ),
sampling_percentage=dict(requires=['sampling_strategy'], ),
sampling_percentage_item_trait=dict(),
sampling_percentage_stage_trait=dict(),
sampling_strategy=dict(requires=['metric_weight'], ),
sampling_strategy_item_trait=dict(),
sampling_strategy_stage_trait=dict(),
save_state=dict(argstr='--save-state %s', ),
shrink_factors=dict(mandatory=True, ),
sigma_units=dict(requires=['smoothing_sigmas'], ),
smoothing_sigmas=dict(mandatory=True, ),
terminal_output=dict(nohash=True, ),
transform_parameters=dict(),
transforms=dict(
argstr='%s',
mandatory=True,
),
use_estimate_learning_rate_once=dict(),
use_histogram_matching=dict(usedefault=True, ),
winsorize_lower_quantile=dict(
argstr='%s',
usedefault=True,
),
winsorize_upper_quantile=dict(
argstr='%s',
usedefault=True,
),
write_composite_transform=dict(
argstr='--write-composite-transform %d',
usedefault=True,
),
)
inputs = Registration.input_spec()
for key, metadata in input_map.items():
for metakey, value in metadata.items():
yield assert_equal, getattr(inputs.traits()[key], metakey), value