liom_toolkit.registration.register module

liom_toolkit.registration.register.align_annotations_to_volume(target_volume: ANTsImage, mask: ANTsImage, template: ANTsImage, data_dir: str, resolution: int = 25, rigid_type: str = 'Similarity', deformable_type: str = 'SyN', keep_intermediary: bool = False, syn_image: dict | None = None, syn_allen: dict | None = None) → ANTsImage

Align an annotation to a target image.

Parameters:

  • target_volume (ants.ANTsImage) – The target image to align to.

  • mask (ants.ANTsImage) – The mask to use in registration.

  • template (ants.ANTsImage) – The template to use for registration.

  • resolution (int) – The resolution of the atlas in micron. Must be 10, 25, 50 or 100 microns

  • data_dir (str) – The directory to use for saving temporary files.

  • rigid_type (str) – The type of rigid registration to use.

  • deformable_type (str) – The type of deformable registration to use.

  • keep_intermediary (bool) – Whether to keep intermediary files or not.

  • syn_image (dict) – The syn transform for the image. If None, it will be calculated.

  • syn_allen (dict) – The syn transform for the Allen template. If None, it will be calculated.

Returns:

The aligned annotation.

Return type:

ants.ANTsImage

liom_toolkit.registration.register.align_brain_region_to_atlas(target_volume: ANTsImage, mask: ANTsImage, template: ANTsImage, region: str, data_dir: str, resolution: int = 25, registration_volume: ANTsImage | None = None, rigid_type: str = 'Similarity', deformable_type: str = 'SyN', keep_intermediary: bool = False, syn_image: dict | None = None, syn_allen: dict | None = None) → ANTsImage

Mask an image with a brain region. Assumes all images are in RAS+ orientation.

Parameters:

  • target_volume (ants.ANTsImage) – The image to mask.

  • mask (ants.ANTsImage) – The mask to use.

  • template (ants.ANTsImage) – The template to use for registration.

  • region (str) – The brain region to use. Will do a lookup in the Allen ontology.

  • data_dir (str) – The directory to use for saving temporary files.

  • resolution (int) – The resolution of the atlas in micron. Must be 10, 25, 50 or 100 microns

  • registration_volume (ants.ANTsImage) – The volume to use for registration. If None, the target_volume will be used.

  • rigid_type (str) – The type of rigid registration to use.

  • deformable_type (str) – The type of deformable registration to use.

  • keep_intermediary (bool) – Whether to write intermediary files or not. Will also save the final masked image.

  • syn_image (dict) – The syn transform for the image. If None, it will be calculated.

  • syn_allen (dict) – The syn transform for the Allen template. If None, it will be calculated.

Returns:

The brain region mask aligned to the target volume.

Return type:

ants.ANTsImage

liom_toolkit.registration.register.align_volume_to_allen(image: ANTsImage, mask: ANTsImage | None, resolution: int = 25) → ANTsImage

Align a volume to the Allen template using the Allen template as a reference.

Parameters:

  • image (ants.ANTsImage) – The image to align

  • mask (ants.ANTsImage | None) – The mask to use in registration

  • resolution (int) – The resolution of the atlas in micron. Must be 10, 25, 50 or 100 microns

Returns:

The aligned image

Return type:

ants.ANTsImage

liom_toolkit.registration.register.deformably_register_volume(image: ~ants.core.ants_image.ANTsImage, mask: ~ants.core.ants_image.ANTsImage | None, template: ~ants.core.ants_image.ANTsImage, rigid_type: str = 'Similarity', deformable_type: str = 'SyN', interpolator: str = 'linear', rigid_interpolator: str = 'linear', use_composite: bool = True) -> (<class 'ants.core.ants_image.ANTsImage'>, <class 'dict'>, <class 'dict'>)

Register an image to a template using a rigid registration followed by a deformable registration.

Parameters:

  • image (ants.ANTsImage) – The image to register

  • mask (ants.ANTsImage) – The mask to use in registration

  • template (ants.ANTsImage) – The template to register to

  • rigid_type (str) – The type of rigid registration to use

  • deformable_type (str) – The type of deformable registration to use

  • interpolator (str) – The interpolator to use to apply the transform.

  • rigid_interpolator (str) – The interpolator to use for applying the rigid transform.

  • use_composite (bool) – Whether to create a composite transform or not

Returns:

The registered image, the transform from the rigid registration, and the transform from the deformable registration

Return type:

tuple[ants.ANTsImage, dict, dict]

liom_toolkit.registration.register.get_transformations_for_atlas(image: ~ants.core.ants_image.ANTsImage, mask: ~ants.core.ants_image.ANTsImage, template: ~ants.core.ants_image.ANTsImage, template_allen: ~ants.core.ants_image.ANTsImage, data_dir: str, rigid_type: str = 'Similarity', deformable_type: str = 'SyN', keep_intermediary: bool = False) -> (<class 'dict'>, <class 'dict'>)

Get the transformations for an image to be aligned to the Allen template.

Parameters:

  • image (ants.ANTsImage) – The image to align.

  • mask (ants.ANTsImage) – The mask of the image to use in registration.

  • template (ants.ANTsImage) – The custom template to use for registration.

  • template_allen (ants.ANTsImage) – The Allen template to use for registration.

  • data_dir (str) – The directory to use for saving temporary files.

  • rigid_type (str) – The type of rigid registration to use.

  • deformable_type (str) – The type of deformable registration to use.

  • keep_intermediary (bool) – Whether to keep intermediary files or not.

Returns:

The transformations for the image to be aligned to the Allen template.

Return type:

tuple[dict, dict]

liom_toolkit.registration.register.rigidly_register_volume(image: ~ants.core.ants_image.ANTsImage, mask: ~ants.core.ants_image.ANTsImage, template: ~ants.core.ants_image.ANTsImage, rigid_type: str = 'Similarity', interpolator: str = 'linear', use_composite: bool = True) -> (<class 'ants.core.ants_image.ANTsImage'>, <class 'dict'>)

Register an image to a template using a rigid registration.

Parameters:

  • image (ants.ANTsImage) – The image to register

  • mask (ants.ANTsImage) – The mask to use in registration

  • template (ants.ANTsImage) – The template to register to

  • rigid_type (str) – The type of rigid registration to use

  • interpolator (str) – The interpolator to use to apply the transform.

  • use_composite (bool) – Whether to create a composite transform or not

Returns:

The registered image and the transform from the rigid registration

Return type:

tuple[ants.ANTsImage, dict]