nipype.interfaces.ants.segmentation module

Wrappers for segmentation utilities within ANTs.

AntsJointFusion

Link to code

alias of JointFusion

Atropos

Link to code

Bases: ANTSCommand

Wrapped executable: Atropos.

A multivariate n-class segmentation algorithm.

A finite mixture modeling (FMM) segmentation approach with possibilities for specifying prior constraints. These prior constraints include the specification of a prior label image, prior probability images (one for each class), and/or an MRF prior to enforce spatial smoothing of the labels. Similar algorithms include FAST and SPM.

Examples

>>> from nipype.interfaces.ants import Atropos
>>> at = Atropos(
...     dimension=3, intensity_images='structural.nii', mask_image='mask.nii',
...     number_of_tissue_classes=2, likelihood_model='Gaussian', save_posteriors=True,
...     mrf_smoothing_factor=0.2, mrf_radius=[1, 1, 1], icm_use_synchronous_update=True,
...     maximum_number_of_icm_terations=1, n_iterations=5, convergence_threshold=0.000001,
...     posterior_formulation='Socrates', use_mixture_model_proportions=True)
>>> at.inputs.initialization = 'Random'
>>> at.cmdline
'Atropos --image-dimensionality 3 --icm [1,1]
--initialization Random[2] --intensity-image structural.nii
--likelihood-model Gaussian --mask-image mask.nii --mrf [0.2,1x1x1] --convergence [5,1e-06]
--output [structural_labeled.nii,POSTERIOR_%02d.nii.gz] --posterior-formulation Socrates[1]
--use-random-seed 1'

>>> at = Atropos(
...     dimension=3, intensity_images='structural.nii', mask_image='mask.nii',
...     number_of_tissue_classes=2, likelihood_model='Gaussian', save_posteriors=True,
...     mrf_smoothing_factor=0.2, mrf_radius=[1, 1, 1], icm_use_synchronous_update=True,
...     maximum_number_of_icm_terations=1, n_iterations=5, convergence_threshold=0.000001,
...     posterior_formulation='Socrates', use_mixture_model_proportions=True)
>>> at.inputs.initialization = 'KMeans'
>>> at.inputs.kmeans_init_centers = [100, 200]
>>> at.cmdline
'Atropos --image-dimensionality 3 --icm [1,1]
--initialization KMeans[2,100,200] --intensity-image structural.nii
--likelihood-model Gaussian --mask-image mask.nii --mrf [0.2,1x1x1] --convergence [5,1e-06]
--output [structural_labeled.nii,POSTERIOR_%02d.nii.gz] --posterior-formulation Socrates[1]
--use-random-seed 1'

>>> at = Atropos(
...     dimension=3, intensity_images='structural.nii', mask_image='mask.nii',
...     number_of_tissue_classes=2, likelihood_model='Gaussian', save_posteriors=True,
...     mrf_smoothing_factor=0.2, mrf_radius=[1, 1, 1], icm_use_synchronous_update=True,
...     maximum_number_of_icm_terations=1, n_iterations=5, convergence_threshold=0.000001,
...     posterior_formulation='Socrates', use_mixture_model_proportions=True)
>>> at.inputs.initialization = 'PriorProbabilityImages'
>>> at.inputs.prior_image = 'BrainSegmentationPrior%02d.nii.gz'
>>> at.inputs.prior_weighting = 0.8
>>> at.inputs.prior_probability_threshold = 0.0000001
>>> at.cmdline
'Atropos --image-dimensionality 3 --icm [1,1]
--initialization PriorProbabilityImages[2,BrainSegmentationPrior%02d.nii.gz,0.8,1e-07]
--intensity-image structural.nii --likelihood-model Gaussian --mask-image mask.nii
--mrf [0.2,1x1x1] --convergence [5,1e-06]
--output [structural_labeled.nii,POSTERIOR_%02d.nii.gz]
--posterior-formulation Socrates[1] --use-random-seed 1'

>>> at = Atropos(
...     dimension=3, intensity_images='structural.nii', mask_image='mask.nii',
...     number_of_tissue_classes=2, likelihood_model='Gaussian', save_posteriors=True,
...     mrf_smoothing_factor=0.2, mrf_radius=[1, 1, 1], icm_use_synchronous_update=True,
...     maximum_number_of_icm_terations=1, n_iterations=5, convergence_threshold=0.000001,
...     posterior_formulation='Socrates', use_mixture_model_proportions=True)
>>> at.inputs.initialization = 'PriorLabelImage'
>>> at.inputs.prior_image = 'segmentation0.nii.gz'
>>> at.inputs.number_of_tissue_classes = 2
>>> at.inputs.prior_weighting = 0.8
>>> at.cmdline
'Atropos --image-dimensionality 3 --icm [1,1]
--initialization PriorLabelImage[2,segmentation0.nii.gz,0.8] --intensity-image structural.nii
--likelihood-model Gaussian --mask-image mask.nii --mrf [0.2,1x1x1] --convergence [5,1e-06]
--output [structural_labeled.nii,POSTERIOR_%02d.nii.gz] --posterior-formulation Socrates[1]
--use-random-seed 1'
Mandatory Inputs:

  • initialization (‘Random’ or ‘Otsu’ or ‘KMeans’ or ‘PriorProbabilityImages’ or ‘PriorLabelImage’) – Maps to a command-line argument: %s. Requires inputs: number_of_tissue_classes.

  • intensity_images (a list of items which are a pathlike object or string representing an existing file) – Maps to a command-line argument: --intensity-image %s....

  • mask_image (a pathlike object or string representing an existing file) – Maps to a command-line argument: --mask-image %s.

  • number_of_tissue_classes (an integer)

Optional Inputs:

  • args (a string) – Additional parameters to the command. Maps to a command-line argument: %s.

  • convergence_threshold (a float) – Requires inputs: n_iterations.

  • dimension (3 or 2 or 4) – Image dimension (2, 3, or 4). Maps to a command-line argument: --image-dimensionality %d. (Nipype default value: 3)

  • environ (a dictionary with keys which are a bytes or None or a value of class ‘str’ and with values which are a bytes or None or a value of class ‘str’) – Environment variables. (Nipype default value: {})

  • icm_use_synchronous_update (a boolean) – Maps to a command-line argument: %s.

  • kmeans_init_centers (a list of at least 1 items which are an integer or a float)

  • likelihood_model (a string) – Maps to a command-line argument: --likelihood-model %s.

  • maximum_number_of_icm_terations (an integer) – Requires inputs: icm_use_synchronous_update.

  • mrf_radius (a list of items which are an integer) – Requires inputs: mrf_smoothing_factor.

  • mrf_smoothing_factor (a float) – Maps to a command-line argument: %s.

  • n_iterations (an integer) – Maps to a command-line argument: %s.

  • num_threads (an integer) – Number of ITK threads to use. (Nipype default value: 1)

  • out_classified_image_name (a pathlike object or string representing a file) – Maps to a command-line argument: %s.

  • output_posteriors_name_template (a string) – (Nipype default value: POSTERIOR_%02d.nii.gz)

  • posterior_formulation (a string) – Maps to a command-line argument: %s.

  • prior_image (a pathlike object or string representing an existing file or a string) – Either a string pattern (e.g., ‘prior%02d.nii’) or an existing vector-image file.

  • prior_probability_threshold (a float) – Requires inputs: prior_weighting.

  • prior_weighting (a float)

  • save_posteriors (a boolean)

  • use_mixture_model_proportions (a boolean) – Requires inputs: posterior_formulation.

  • use_random_seed (a boolean) – Use random seed value over constant. Maps to a command-line argument: --use-random-seed %d. (Nipype default value: True)

Outputs:

  • classified_image (a pathlike object or string representing an existing file)

  • posteriors (a list of items which are a pathlike object or string representing a file)

BrainExtraction

Link to code

Bases: ANTSCommand

Wrapped executable: antsBrainExtraction.sh.

Atlas-based brain extraction.

Examples

>>> from nipype.interfaces.ants.segmentation import BrainExtraction
>>> brainextraction = BrainExtraction()
>>> brainextraction.inputs.dimension = 3
>>> brainextraction.inputs.anatomical_image ='T1.nii.gz'
>>> brainextraction.inputs.brain_template = 'study_template.nii.gz'
>>> brainextraction.inputs.brain_probability_mask ='ProbabilityMaskOfStudyTemplate.nii.gz'
>>> brainextraction.cmdline
'antsBrainExtraction.sh -a T1.nii.gz -m ProbabilityMaskOfStudyTemplate.nii.gz
-e study_template.nii.gz -d 3 -s nii.gz -o highres001_'
Mandatory Inputs:

  • anatomical_image (a pathlike object or string representing an existing file) – Structural image, typically T1. If more than one anatomical image is specified, subsequently specified images are used during the segmentation process. However, only the first image is used in the registration of priors. Our suggestion would be to specify the T1 as the first image. Anatomical template created using e.g. LPBA40 data set with buildtemplateparallel.sh in ANTs. Maps to a command-line argument: -a %s.

  • brain_probability_mask (a pathlike object or string representing an existing file) – Brain probability mask created using e.g. LPBA40 data set which have brain masks defined, and warped to anatomical template and averaged resulting in a probability image. Maps to a command-line argument: -m %s.

  • brain_template (a pathlike object or string representing an existing file) – Anatomical template created using e.g. LPBA40 data set with buildtemplateparallel.sh in ANTs. Maps to a command-line argument: -e %s.

Optional Inputs:

  • args (a string) – Additional parameters to the command. Maps to a command-line argument: %s.

  • debug (a boolean) – If > 0, runs a faster version of the script. Only for testing. Implies -u 0. Requires single thread computation for complete reproducibility. Maps to a command-line argument: -z 1.

  • dimension (3 or 2) – Image dimension (2 or 3). Maps to a command-line argument: -d %d. (Nipype default value: 3)

  • environ (a dictionary with keys which are a bytes or None or a value of class ‘str’ and with values which are a bytes or None or a value of class ‘str’) – Environment variables. (Nipype default value: {})

  • extraction_registration_mask (a pathlike object or string representing an existing file) – Mask (defined in the template space) used during registration for brain extraction. To limit the metric computation to a specific region. Maps to a command-line argument: -f %s.

  • image_suffix (a string) – Any of standard ITK formats, nii.gz is default. Maps to a command-line argument: -s %s. (Nipype default value: nii.gz)

  • keep_temporary_files (an integer) – Keep brain extraction/segmentation warps, etc (default = 0). Maps to a command-line argument: -k %d.

  • num_threads (an integer) – Number of ITK threads to use. (Nipype default value: 1)

  • out_prefix (a string) – Prefix that is prepended to all output files. Maps to a command-line argument: -o %s. (Nipype default value: highres001_)

  • use_floatingpoint_precision (0 or 1) – Use floating point precision in registrations (default = 0). Maps to a command-line argument: -q %d.

  • use_random_seeding (0 or 1) – Use random number generated from system clock in Atropos (default = 1). Maps to a command-line argument: -u %d.

Outputs:

  • BrainExtractionBrain (a pathlike object or string representing an existing file) – Brain extraction image.

  • BrainExtractionCSF (a pathlike object or string representing an existing file) – Segmentation mask with only CSF.

  • BrainExtractionGM (a pathlike object or string representing an existing file) – Segmentation mask with only grey matter.

  • BrainExtractionInitialAffine (a pathlike object or string representing an existing file)

  • BrainExtractionInitialAffineFixed (a pathlike object or string representing an existing file)

  • BrainExtractionInitialAffineMoving (a pathlike object or string representing an existing file)

  • BrainExtractionLaplacian (a pathlike object or string representing an existing file)

  • BrainExtractionMask (a pathlike object or string representing an existing file) – Brain extraction mask.

  • BrainExtractionPrior0GenericAffine (a pathlike object or string representing an existing file)

  • BrainExtractionPrior1InverseWarp (a pathlike object or string representing an existing file)

  • BrainExtractionPrior1Warp (a pathlike object or string representing an existing file)

  • BrainExtractionPriorWarped (a pathlike object or string representing an existing file)

  • BrainExtractionSegmentation (a pathlike object or string representing an existing file) – Segmentation mask with CSF, GM, and WM.

  • BrainExtractionTemplateLaplacian (a pathlike object or string representing an existing file)

  • BrainExtractionTmp (a pathlike object or string representing an existing file)

  • BrainExtractionWM (a pathlike object or string representing an existing file) – Segmenration mask with only white matter.

  • N4Corrected0 (a pathlike object or string representing an existing file) – N4 bias field corrected image.

  • N4Truncated0 (a pathlike object or string representing an existing file)

CorticalThickness

Link to code

Bases: ANTSCommand

Wrapped executable: antsCorticalThickness.sh.

Examples

>>> from nipype.interfaces.ants.segmentation import CorticalThickness
>>> corticalthickness = CorticalThickness()
>>> corticalthickness.inputs.dimension = 3
>>> corticalthickness.inputs.anatomical_image ='T1.nii.gz'
>>> corticalthickness.inputs.brain_template = 'study_template.nii.gz'
>>> corticalthickness.inputs.brain_probability_mask ='ProbabilityMaskOfStudyTemplate.nii.gz'
>>> corticalthickness.inputs.segmentation_priors = ['BrainSegmentationPrior01.nii.gz',
...                                                 'BrainSegmentationPrior02.nii.gz',
...                                                 'BrainSegmentationPrior03.nii.gz',
...                                                 'BrainSegmentationPrior04.nii.gz']
>>> corticalthickness.inputs.t1_registration_template = 'brain_study_template.nii.gz'
>>> corticalthickness.cmdline
'antsCorticalThickness.sh -a T1.nii.gz -m ProbabilityMaskOfStudyTemplate.nii.gz
-e study_template.nii.gz -d 3 -s nii.gz -o antsCT_
-p nipype_priors/BrainSegmentationPrior%02d.nii.gz -t brain_study_template.nii.gz'
Mandatory Inputs:

  • anatomical_image (a pathlike object or string representing an existing file) – Structural intensity image, typically T1. If more than one anatomical image is specified, subsequently specified images are used during the segmentation process. However, only the first image is used in the registration of priors. Our suggestion would be to specify the T1 as the first image. Maps to a command-line argument: -a %s.

  • brain_probability_mask (a pathlike object or string representing an existing file) – Brain probability mask in template space. Maps to a command-line argument: -m %s.

  • brain_template (a pathlike object or string representing an existing file) – Anatomical intensity template (possibly created using a population data set with buildtemplateparallel.sh in ANTs). This template is not skull-stripped. Maps to a command-line argument: -e %s.

  • segmentation_priors (a list of items which are a pathlike object or string representing an existing file) – Maps to a command-line argument: -p %s.

  • t1_registration_template (a pathlike object or string representing an existing file) – Anatomical intensity template (assumed to be skull-stripped). A common case would be where this would be the same template as specified in the -e option which is not skull stripped. Maps to a command-line argument: -t %s.

Optional Inputs:

  • args (a string) – Additional parameters to the command. Maps to a command-line argument: %s.

  • b_spline_smoothing (a boolean) – Use B-spline SyN for registrations and B-spline exponential mapping in DiReCT. Maps to a command-line argument: -v.

  • cortical_label_image (a pathlike object or string representing an existing file) – Cortical ROI labels to use as a prior for ATITH.

  • debug (a boolean) – If > 0, runs a faster version of the script. Only for testing. Implies -u 0. Requires single thread computation for complete reproducibility. Maps to a command-line argument: -z 1.

  • dimension (3 or 2) – Image dimension (2 or 3). Maps to a command-line argument: -d %d. (Nipype default value: 3)

  • environ (a dictionary with keys which are a bytes or None or a value of class ‘str’ and with values which are a bytes or None or a value of class ‘str’) – Environment variables. (Nipype default value: {})

  • extraction_registration_mask (a pathlike object or string representing an existing file) – Mask (defined in the template space) used during registration for brain extraction. Maps to a command-line argument: -f %s.

  • image_suffix (a string) – Any of standard ITK formats, nii.gz is default. Maps to a command-line argument: -s %s. (Nipype default value: nii.gz)

  • keep_temporary_files (an integer) – Keep brain extraction/segmentation warps, etc (default = 0). Maps to a command-line argument: -k %d.

  • label_propagation (a string) – Incorporate a distance prior one the posterior formulation. Should be of the form ‘label[lambda,boundaryProbability]’ where label is a value of 1,2,3,… denoting label ID. The label probability for anything outside the current label = boundaryProbability * exp( -lambda * distanceFromBoundary ) Intuitively, smaller lambda values will increase the spatial capture range of the distance prior. To apply to all label values, simply omit specifying the label, i.e. -l [lambda,boundaryProbability]. Maps to a command-line argument: -l %s.

  • max_iterations (an integer) – ANTS registration max iterations (default = 100x100x70x20). Maps to a command-line argument: -i %d.

  • num_threads (an integer) – Number of ITK threads to use. (Nipype default value: 1)

  • out_prefix (a string) – Prefix that is prepended to all output files. Maps to a command-line argument: -o %s. (Nipype default value: antsCT_)

  • posterior_formulation (a string) – Atropos posterior formulation and whether or not to use mixture model proportions. e.g ‘Socrates[1]’ (default) or ‘Aristotle[1]’. Choose the latter if you want use the distance priors (see also the -l option for label propagation control). Maps to a command-line argument: -b %s.

  • prior_segmentation_weight (a float) – Atropos spatial prior probability weight for the segmentation. Maps to a command-line argument: -w %f.

  • quick_registration (a boolean) – If = 1, use antsRegistrationSyNQuick.sh as the basis for registration during brain extraction, brain segmentation, and (optional) normalization to a template. Otherwise use antsRegistrationSyN.sh (default = 0). Maps to a command-line argument: -q 1.

  • segmentation_iterations (an integer) – N4 -> Atropos -> N4 iterations during segmentation (default = 3). Maps to a command-line argument: -n %d.

  • use_floatingpoint_precision (0 or 1) – Use floating point precision in registrations (default = 0). Maps to a command-line argument: -j %d.

  • use_random_seeding (0 or 1) – Use random number generated from system clock in Atropos (default = 1). Maps to a command-line argument: -u %d.

Outputs:

  • BrainExtractionMask (a pathlike object or string representing an existing file) – Brain extraction mask.

  • BrainSegmentation (a pathlike object or string representing an existing file) – Brain segmentation image.

  • BrainSegmentationN4 (a pathlike object or string representing an existing file) – N4 corrected image.

  • BrainSegmentationPosteriors (a list of items which are a pathlike object or string representing an existing file) – Posterior probability images.

  • BrainVolumes (a pathlike object or string representing an existing file) – Brain volumes as text.

  • CorticalThickness (a pathlike object or string representing an existing file) – Cortical thickness file.

  • CorticalThicknessNormedToTemplate (a pathlike object or string representing an existing file) – Normalized cortical thickness.

  • ExtractedBrainN4 (a pathlike object or string representing an existing file) – Extracted brain from N4 image.

  • SubjectToTemplate0GenericAffine (a pathlike object or string representing an existing file) – Template to subject inverse affine.

  • SubjectToTemplate1Warp (a pathlike object or string representing an existing file) – Template to subject inverse warp.

  • SubjectToTemplateLogJacobian (a pathlike object or string representing an existing file) – Template to subject log jacobian.

  • TemplateToSubject0Warp (a pathlike object or string representing an existing file) – Template to subject warp.

  • TemplateToSubject1GenericAffine (a pathlike object or string representing an existing file) – Template to subject affine.

DenoiseImage

Link to code

Bases: ANTSCommand

Wrapped executable: DenoiseImage.

Examples

>>> import copy
>>> from nipype.interfaces.ants import DenoiseImage
>>> denoise = DenoiseImage()
>>> denoise.inputs.dimension = 3
>>> denoise.inputs.input_image = 'im1.nii'
>>> denoise.cmdline
'DenoiseImage -d 3 -i im1.nii -n Gaussian -o im1_noise_corrected.nii -s 1'

>>> denoise_2 = copy.deepcopy(denoise)
>>> denoise_2.inputs.output_image = 'output_corrected_image.nii.gz'
>>> denoise_2.inputs.noise_model = 'Rician'
>>> denoise_2.inputs.shrink_factor = 2
>>> denoise_2.cmdline
'DenoiseImage -d 3 -i im1.nii -n Rician -o output_corrected_image.nii.gz -s 2'

>>> denoise_3 = DenoiseImage()
>>> denoise_3.inputs.input_image = 'im1.nii'
>>> denoise_3.inputs.save_noise = True
>>> denoise_3.cmdline
'DenoiseImage -i im1.nii -n Gaussian -o [ im1_noise_corrected.nii, im1_noise.nii ] -s 1'
Mandatory Inputs:

  • input_image (a pathlike object or string representing an existing file) – A scalar image is expected as input for noise correction. Maps to a command-line argument: -i %s.

  • save_noise (a boolean) – True if the estimated noise should be saved to file. Mutually exclusive with inputs: noise_image. (Nipype default value: False)

Optional Inputs:

  • args (a string) – Additional parameters to the command. Maps to a command-line argument: %s.

  • dimension (2 or 3 or 4) – This option forces the image to be treated as a specified-dimensional image. If not specified, the program tries to infer the dimensionality from the input image. Maps to a command-line argument: -d %d.

  • environ (a dictionary with keys which are a bytes or None or a value of class ‘str’ and with values which are a bytes or None or a value of class ‘str’) – Environment variables. (Nipype default value: {})

  • noise_image (a pathlike object or string representing a file) – Filename for the estimated noise.

  • noise_model (‘Gaussian’ or ‘Rician’) – Employ a Rician or Gaussian noise model. Maps to a command-line argument: -n %s. (Nipype default value: Gaussian)

  • num_threads (an integer) – Number of ITK threads to use. (Nipype default value: 1)

  • output_image (a pathlike object or string representing a file) – The output consists of the noise corrected version of the input image. Maps to a command-line argument: -o %s.

  • shrink_factor (an integer) – Running noise correction on large images can be time consuming. To lessen computation time, the input image can be resampled. The shrink factor, specified as a single integer, describes this resampling. Shrink factor = 1 is the default. Maps to a command-line argument: -s %s. (Nipype default value: 1)

  • verbose (a boolean) – Verbose output. Maps to a command-line argument: -v.

Outputs:

  • noise_image (a pathlike object or string representing a file)

  • output_image (a pathlike object or string representing an existing file)

JointFusion

Link to code

Bases: ANTSCommand

Wrapped executable: antsJointFusion.

An image fusion algorithm.

Developed by Hongzhi Wang and Paul Yushkevich, and it won segmentation challenges at MICCAI 2012 and MICCAI 2013. The original label fusion framework was extended to accommodate intensities by Brian Avants. This implementation is based on Paul’s original ITK-style implementation and Brian’s ANTsR implementation.

References include 1) H. Wang, J. W. Suh, S. Das, J. Pluta, C. Craige, P. Yushkevich, Multi-atlas segmentation with joint label fusion IEEE Trans. on Pattern Analysis and Machine Intelligence, 35(3), 611-623, 2013. and 2) H. Wang and P. A. Yushkevich, Multi-atlas segmentation with joint label fusion and corrective learning–an open source implementation, Front. Neuroinform., 2013.

Examples

>>> from nipype.interfaces.ants import JointFusion
>>> jf = JointFusion()
>>> jf.inputs.out_label_fusion = 'ants_fusion_label_output.nii'
>>> jf.inputs.atlas_image = [ ['rc1s1.nii','rc1s2.nii'] ]
>>> jf.inputs.atlas_segmentation_image = ['segmentation0.nii.gz']
>>> jf.inputs.target_image = ['im1.nii']
>>> jf.cmdline
"antsJointFusion -a 0.1 -g ['rc1s1.nii', 'rc1s2.nii'] -l segmentation0.nii.gz
-b 2.0 -o ants_fusion_label_output.nii -s 3x3x3 -t ['im1.nii']"

>>> jf.inputs.target_image = [ ['im1.nii', 'im2.nii'] ]
>>> jf.cmdline
"antsJointFusion -a 0.1 -g ['rc1s1.nii', 'rc1s2.nii'] -l segmentation0.nii.gz
-b 2.0 -o ants_fusion_label_output.nii -s 3x3x3 -t ['im1.nii', 'im2.nii']"

>>> jf.inputs.atlas_image = [ ['rc1s1.nii','rc1s2.nii'],
...                                        ['rc2s1.nii','rc2s2.nii'] ]
>>> jf.inputs.atlas_segmentation_image = ['segmentation0.nii.gz',
...                                                    'segmentation1.nii.gz']
>>> jf.cmdline
"antsJointFusion -a 0.1 -g ['rc1s1.nii', 'rc1s2.nii'] -g ['rc2s1.nii', 'rc2s2.nii']
-l segmentation0.nii.gz -l segmentation1.nii.gz -b 2.0 -o ants_fusion_label_output.nii
-s 3x3x3 -t ['im1.nii', 'im2.nii']"

>>> jf.inputs.dimension = 3
>>> jf.inputs.alpha = 0.5
>>> jf.inputs.beta = 1.0
>>> jf.inputs.patch_radius = [3,2,1]
>>> jf.inputs.search_radius = [3]
>>> jf.cmdline
"antsJointFusion -a 0.5 -g ['rc1s1.nii', 'rc1s2.nii'] -g ['rc2s1.nii', 'rc2s2.nii']
-l segmentation0.nii.gz -l segmentation1.nii.gz -b 1.0 -d 3 -o ants_fusion_label_output.nii
-p 3x2x1 -s 3 -t ['im1.nii', 'im2.nii']"

>>> jf.inputs.search_radius = ['mask.nii']
>>> jf.inputs.verbose = True
>>> jf.inputs.exclusion_image = ['roi01.nii', 'roi02.nii']
>>> jf.inputs.exclusion_image_label = ['1','2']
>>> jf.cmdline
"antsJointFusion -a 0.5 -g ['rc1s1.nii', 'rc1s2.nii'] -g ['rc2s1.nii', 'rc2s2.nii']
-l segmentation0.nii.gz -l segmentation1.nii.gz -b 1.0 -d 3 -e 1[roi01.nii] -e 2[roi02.nii]
-o ants_fusion_label_output.nii -p 3x2x1 -s mask.nii -t ['im1.nii', 'im2.nii'] -v"

>>> jf.inputs.out_label_fusion = 'ants_fusion_label_output.nii'
>>> jf.inputs.out_intensity_fusion_name_format = 'ants_joint_fusion_intensity_%d.nii.gz'
>>> jf.inputs.out_label_post_prob_name_format = 'ants_joint_fusion_posterior_%d.nii.gz'
>>> jf.inputs.out_atlas_voting_weight_name_format = 'ants_joint_fusion_voting_weight_%d.nii.gz'
>>> jf.cmdline
"antsJointFusion -a 0.5 -g ['rc1s1.nii', 'rc1s2.nii'] -g ['rc2s1.nii', 'rc2s2.nii']
-l segmentation0.nii.gz -l segmentation1.nii.gz -b 1.0 -d 3 -e 1[roi01.nii] -e 2[roi02.nii]
-o [ants_fusion_label_output.nii, ants_joint_fusion_intensity_%d.nii.gz,
ants_joint_fusion_posterior_%d.nii.gz, ants_joint_fusion_voting_weight_%d.nii.gz]
-p 3x2x1 -s mask.nii -t ['im1.nii', 'im2.nii'] -v"
Mandatory Inputs:

  • atlas_image (a list of items which are a list of items which are a pathlike object or string representing an existing file) – The atlas image (or multimodal atlas images) assumed to be aligned to a common image domain. Maps to a command-line argument: -g %s....

  • atlas_segmentation_image (a list of items which are a pathlike object or string representing an existing file) – The atlas segmentation images. For performing label fusion the number of specified segmentations should be identical to the number of atlas image sets. Maps to a command-line argument: -l %s....

  • target_image (a list of items which are a list of items which are a pathlike object or string representing an existing file) – The target image (or multimodal target images) assumed to be aligned to a common image domain. Maps to a command-line argument: -t %s.

Optional Inputs:

  • alpha (a float) – Regularization term added to matrix Mx for calculating the inverse. Default = 0.1. Maps to a command-line argument: -a %s. (Nipype default value: 0.1)

  • args (a string) – Additional parameters to the command. Maps to a command-line argument: %s.

  • beta (a float) – Exponent for mapping intensity difference to the joint error. Default = 2.0. Maps to a command-line argument: -b %s. (Nipype default value: 2.0)

  • constrain_nonnegative (a boolean) – Constrain solution to non-negative weights. Maps to a command-line argument: -c. (Nipype default value: False)

  • dimension (3 or 2 or 4) – This option forces the image to be treated as a specified-dimensional image. If not specified, the program tries to infer the dimensionality from the input image. Maps to a command-line argument: -d %d.

  • environ (a dictionary with keys which are a bytes or None or a value of class ‘str’ and with values which are a bytes or None or a value of class ‘str’) – Environment variables. (Nipype default value: {})

  • exclusion_image (a list of items which are a pathlike object or string representing an existing file) – Specify an exclusion region for the given label.

  • exclusion_image_label (a list of items which are a string) – Specify a label for the exclusion region. Maps to a command-line argument: -e %s. Requires inputs: exclusion_image.

  • mask_image (a pathlike object or string representing an existing file) – If a mask image is specified, fusion is only performed in the mask region. Maps to a command-line argument: -x %s.

  • num_threads (an integer) – Number of ITK threads to use. (Nipype default value: 1)

  • out_atlas_voting_weight_name_format (a string) – Optional atlas voting weight image file name format. Requires inputs: out_label_fusion, out_intensity_fusion_name_format, out_label_post_prob_name_format.

  • out_intensity_fusion_name_format (a string) – Optional intensity fusion image file name format. (e.g. “antsJointFusionIntensity_%d.nii.gz”).

  • out_label_fusion (a pathlike object or string representing a file) – The output label fusion image. Maps to a command-line argument: %s.

  • out_label_post_prob_name_format (a string) – Optional label posterior probability image file name format. Requires inputs: out_label_fusion, out_intensity_fusion_name_format.

  • patch_metric (‘PC’ or ‘MSQ’) – Metric to be used in determining the most similar neighborhood patch. Options include Pearson’s correlation (PC) and mean squares (MSQ). Default = PC (Pearson correlation). Maps to a command-line argument: -m %s.

  • patch_radius (a list of items which are a value of class ‘int’) – Patch radius for similarity measures. Default: 2x2x2. Maps to a command-line argument: -p %s.

  • retain_atlas_voting_images (a boolean) – Retain atlas voting images. Default = false. Maps to a command-line argument: -f. (Nipype default value: False)

  • retain_label_posterior_images (a boolean) – Retain label posterior probability images. Requires atlas segmentations to be specified. Default = false. Maps to a command-line argument: -r. Requires inputs: atlas_segmentation_image. (Nipype default value: False)

  • search_radius (a list of from 1 to 3 items which are any value) – Search radius for similarity measures. Default = 3x3x3. One can also specify an image where the value at the voxel specifies the isotropic search radius at that voxel. Maps to a command-line argument: -s %s. (Nipype default value: [3, 3, 3])

  • verbose (a boolean) – Verbose output. Maps to a command-line argument: -v.

Outputs:

  • out_atlas_voting_weight (a list of items which are a pathlike object or string representing an existing file)

  • out_intensity_fusion (a list of items which are a pathlike object or string representing an existing file)

  • out_label_fusion (a pathlike object or string representing an existing file)

  • out_label_post_prob (a list of items which are a pathlike object or string representing an existing file)

KellyKapowski

Link to code

Bases: ANTSCommand

Wrapped executable: KellyKapowski.

Nipype Interface to ANTs’ KellyKapowski, also known as DiReCT.

DiReCT is a registration based estimate of cortical thickness. It was published in S. R. Das, B. B. Avants, M. Grossman, and J. C. Gee, Registration based cortical thickness measurement, Neuroimage 2009, 45:867–879.

Examples

>>> from nipype.interfaces.ants.segmentation import KellyKapowski
>>> kk = KellyKapowski()
>>> kk.inputs.dimension = 3
>>> kk.inputs.segmentation_image = "segmentation0.nii.gz"
>>> kk.inputs.convergence = "[45,0.0,10]"
>>> kk.inputs.thickness_prior_estimate = 10
>>> kk.cmdline
'KellyKapowski --convergence "[45,0.0,10]"
--output "[segmentation0_cortical_thickness.nii.gz,segmentation0_warped_white_matter.nii.gz]"
--image-dimensionality 3 --gradient-step 0.025000
--maximum-number-of-invert-displacement-field-iterations 20 --number-of-integration-points 10
--segmentation-image "[segmentation0.nii.gz,2,3]" --smoothing-variance 1.000000
--smoothing-velocity-field-parameter 1.500000 --thickness-prior-estimate 10.000000'
Mandatory Inputs:

segmentation_image (a pathlike object or string representing an existing file) – A segmentation image must be supplied labeling the gray and white matters. Default values = 2 and 3, respectively. Maps to a command-line argument: --segmentation-image "%s".

Optional Inputs:

  • args (a string) – Additional parameters to the command. Maps to a command-line argument: %s.

  • convergence (a string) – Convergence is determined by fitting a line to the normalized energy profile of the last N iterations (where N is specified by the window size) and determining the slope which is then compared with the convergence threshold. Maps to a command-line argument: --convergence "%s". (Nipype default value: [50,0.001,10])

  • cortical_thickness (a pathlike object or string representing a file) – Filename for the cortical thickness. Maps to a command-line argument: --output "%s".

  • dimension (3 or 2) – Image dimension (2 or 3). Maps to a command-line argument: --image-dimensionality %d. (Nipype default value: 3)

  • environ (a dictionary with keys which are a bytes or None or a value of class ‘str’ and with values which are a bytes or None or a value of class ‘str’) – Environment variables. (Nipype default value: {})

  • gradient_step (a float) – Gradient step size for the optimization. Maps to a command-line argument: --gradient-step %f. (Nipype default value: 0.025)

  • gray_matter_label (an integer) – The label value for the gray matter label in the segmentation_image. (Nipype default value: 2)

  • gray_matter_prob_image (a pathlike object or string representing an existing file) – In addition to the segmentation image, a gray matter probability image can be used. If no such image is supplied, one is created using the segmentation image and a variance of 1.0 mm. Maps to a command-line argument: --gray-matter-probability-image "%s".

  • max_invert_displacement_field_iters (an integer) – Maximum number of iterations for estimating the invertdisplacement field. Maps to a command-line argument: --maximum-number-of-invert-displacement-field-iterations %d. (Nipype default value: 20)

  • num_threads (an integer) – Number of ITK threads to use. (Nipype default value: 1)

  • number_integration_points (an integer) – Number of compositions of the diffeomorphism per iteration. Maps to a command-line argument: --number-of-integration-points %d. (Nipype default value: 10)

  • smoothing_variance (a float) – Defines the Gaussian smoothing of the hit and total images. Maps to a command-line argument: --smoothing-variance %f. (Nipype default value: 1.0)

  • smoothing_velocity_field (a float) – Defines the Gaussian smoothing of the velocity field (default = 1.5). If the b-spline smoothing option is chosen, then this defines the isotropic mesh spacing for the smoothing spline (default = 15). Maps to a command-line argument: --smoothing-velocity-field-parameter %f. (Nipype default value: 1.5)

  • thickness_prior_estimate (a float) – Provides a prior constraint on the final thickness measurement in mm. Maps to a command-line argument: --thickness-prior-estimate %f. (Nipype default value: 10)

  • thickness_prior_image (a pathlike object or string representing an existing file) – An image containing spatially varying prior thickness values. Maps to a command-line argument: --thickness-prior-image "%s".

  • use_bspline_smoothing (a boolean) – Sets the option for B-spline smoothing of the velocity field. Maps to a command-line argument: --use-bspline-smoothing 1.

  • warped_white_matter (a pathlike object or string representing a file) – Filename for the warped white matter file.

  • white_matter_label (an integer) – The label value for the white matter label in the segmentation_image. (Nipype default value: 3)

  • white_matter_prob_image (a pathlike object or string representing an existing file) – In addition to the segmentation image, a white matter probability image can be used. If no such image is supplied, one is created using the segmentation image and a variance of 1.0 mm. Maps to a command-line argument: --white-matter-probability-image "%s".

Outputs:

  • cortical_thickness (a pathlike object or string representing a file) – A thickness map defined in the segmented gray matter.

  • warped_white_matter (a pathlike object or string representing a file) – A warped white matter image.

LaplacianThickness

Link to code

Bases: ANTSCommand

Wrapped executable: LaplacianThickness.

Calculates the cortical thickness from an anatomical image

Examples

>>> from nipype.interfaces.ants import LaplacianThickness
>>> cort_thick = LaplacianThickness()
>>> cort_thick.inputs.input_wm = 'white_matter.nii.gz'
>>> cort_thick.inputs.input_gm = 'gray_matter.nii.gz'
>>> cort_thick.cmdline
'LaplacianThickness white_matter.nii.gz gray_matter.nii.gz white_matter_thickness.nii.gz'

>>> cort_thick.inputs.output_image = 'output_thickness.nii.gz'
>>> cort_thick.cmdline
'LaplacianThickness white_matter.nii.gz gray_matter.nii.gz output_thickness.nii.gz'
Mandatory Inputs:

  • input_gm (a pathlike object or string representing a file) – Gray matter segmentation image. Maps to a command-line argument: %s (position: 2).

  • input_wm (a pathlike object or string representing a file) – White matter segmentation image. Maps to a command-line argument: %s (position: 1).

Optional Inputs:

  • args (a string) – Additional parameters to the command. Maps to a command-line argument: %s.

  • dT (a float) – Time delta used during integration (defaults to 0.01). Maps to a command-line argument: %s (position: 6). Requires inputs: prior_thickness.

  • environ (a dictionary with keys which are a bytes or None or a value of class ‘str’ and with values which are a bytes or None or a value of class ‘str’) – Environment variables. (Nipype default value: {})

  • num_threads (an integer) – Number of ITK threads to use. (Nipype default value: 1)

  • output_image (a string) – Name of output file. Maps to a command-line argument: %s (position: 3).

  • prior_thickness (a float) – Prior thickness (defaults to 500). Maps to a command-line argument: %s (position: 5). Requires inputs: smooth_param.

  • smooth_param (a float) – Sigma of the Laplacian Recursive Image Filter (defaults to 1). Maps to a command-line argument: %s (position: 4).

  • sulcus_prior (a float) – Positive floating point number for sulcus prior. Authors said that 0.15 might be a reasonable value. Maps to a command-line argument: %s (position: 7). Requires inputs: dT.

  • tolerance (a float) – Tolerance to reach during optimization (defaults to 0.001). Maps to a command-line argument: %s (position: 8). Requires inputs: sulcus_prior.

Outputs:

output_image (a pathlike object or string representing an existing file) – Cortical thickness.

N4BiasFieldCorrection

Link to code

Bases: ANTSCommand, CopyHeaderInterface

Wrapped executable: N4BiasFieldCorrection.

Bias field correction.

N4 is a variant of the popular N3 (nonparameteric nonuniform normalization) retrospective bias correction algorithm. Based on the assumption that the corruption of the low frequency bias field can be modeled as a convolution of the intensity histogram by a Gaussian, the basic algorithmic protocol is to iterate between deconvolving the intensity histogram by a Gaussian, remapping the intensities, and then spatially smoothing this result by a B-spline modeling of the bias field itself. The modifications from and improvements obtained over the original N3 algorithm are described in [Tustison2010].

[Tustison2010]

N. Tustison et al., N4ITK: Improved N3 Bias Correction, IEEE Transactions on Medical Imaging, 29(6):1310-1320, June 2010.

Examples

>>> import copy
>>> from nipype.interfaces.ants import N4BiasFieldCorrection
>>> n4 = N4BiasFieldCorrection()
>>> n4.inputs.dimension = 3
>>> n4.inputs.input_image = 'structural.nii'
>>> n4.inputs.bspline_fitting_distance = 300
>>> n4.inputs.shrink_factor = 3
>>> n4.inputs.n_iterations = [50,50,30,20]
>>> n4.cmdline
'N4BiasFieldCorrection --bspline-fitting [ 300 ]
-d 3 --input-image structural.nii
--convergence [ 50x50x30x20 ] --output structural_corrected.nii
--shrink-factor 3'

>>> n4_2 = copy.deepcopy(n4)
>>> n4_2.inputs.convergence_threshold = 1e-6
>>> n4_2.cmdline
'N4BiasFieldCorrection --bspline-fitting [ 300 ]
-d 3 --input-image structural.nii
--convergence [ 50x50x30x20, 1e-06 ] --output structural_corrected.nii
--shrink-factor 3'

>>> n4_3 = copy.deepcopy(n4_2)
>>> n4_3.inputs.bspline_order = 5
>>> n4_3.cmdline
'N4BiasFieldCorrection --bspline-fitting [ 300, 5 ]
-d 3 --input-image structural.nii
--convergence [ 50x50x30x20, 1e-06 ] --output structural_corrected.nii
--shrink-factor 3'

>>> n4_4 = N4BiasFieldCorrection()
>>> n4_4.inputs.input_image = 'structural.nii'
>>> n4_4.inputs.save_bias = True
>>> n4_4.inputs.dimension = 3
>>> n4_4.cmdline
'N4BiasFieldCorrection -d 3 --input-image structural.nii
--output [ structural_corrected.nii, structural_bias.nii ]'

>>> n4_5 = N4BiasFieldCorrection()
>>> n4_5.inputs.input_image = 'structural.nii'
>>> n4_5.inputs.dimension = 3
>>> n4_5.inputs.histogram_sharpening = (0.12, 0.02, 200)
>>> n4_5.cmdline
'N4BiasFieldCorrection -d 3  --histogram-sharpening [0.12,0.02,200]
--input-image structural.nii --output structural_corrected.nii'
Mandatory Inputs:

  • copy_header (a boolean) – Copy headers of the original image into the output (corrected) file. (Nipype default value: False)

  • input_image (a pathlike object or string representing a file) – Input for bias correction. Negative values or values close to zero should be processed prior to correction. Maps to a command-line argument: --input-image %s.

  • save_bias (a boolean) – True if the estimated bias should be saved to file. Mutually exclusive with inputs: bias_image. (Nipype default value: False)

Optional Inputs:

  • args (a string) – Additional parameters to the command. Maps to a command-line argument: %s.

  • bias_image (a pathlike object or string representing a file) – Filename for the estimated bias.

  • bspline_fitting_distance (a float) – Maps to a command-line argument: --bspline-fitting %s.

  • bspline_order (an integer) – Requires inputs: bspline_fitting_distance.

  • convergence_threshold (a float) – Requires inputs: n_iterations.

  • dimension (3 or 2 or 4) – Image dimension (2, 3 or 4). Maps to a command-line argument: -d %d. (Nipype default value: 3)

  • environ (a dictionary with keys which are a bytes or None or a value of class ‘str’ and with values which are a bytes or None or a value of class ‘str’) – Environment variables. (Nipype default value: {})

  • histogram_sharpening (a tuple of the form: (a float, a float, an integer)) – Three-values tuple of histogram sharpening parameters (FWHM, wienerNose, numberOfHistogramBins). These options describe the histogram sharpening parameters, i.e. the deconvolution step parameters described in the original N3 algorithm. The default values have been shown to work fairly well. Maps to a command-line argument: --histogram-sharpening [%g,%g,%d].

  • mask_image (a pathlike object or string representing a file) – Image to specify region to perform final bias correction in. Maps to a command-line argument: --mask-image %s.

  • n_iterations (a list of items which are an integer) – Maps to a command-line argument: --convergence %s.

  • num_threads (an integer) – Number of ITK threads to use. (Nipype default value: 1)

  • output_image (a string) – Output file name. Maps to a command-line argument: --output %s.

  • rescale_intensities (a boolean) – [NOTE: Only ANTs>=2.1.0] At each iteration, a new intensity mapping is calculated and applied but there is nothing which constrains the new intensity range to be within certain values. The result is that the range can “drift” from the original at each iteration. This option rescales to the [min,max] range of the original image intensities within the user-specified mask. Maps to a command-line argument: -r. (Nipype default value: False)

  • shrink_factor (an integer) – Maps to a command-line argument: --shrink-factor %d.

  • weight_image (a pathlike object or string representing a file) – Image for relative weighting (e.g. probability map of the white matter) of voxels during the B-spline fitting. . Maps to a command-line argument: --weight-image %s.

Outputs:

  • bias_image (a pathlike object or string representing an existing file) – Estimated bias.

  • output_image (a pathlike object or string representing an existing file) – Warped image.