registration

registration.registrator

class brainles_preprocessing.registration.registrator.Registrator

Bases: ABC

_abc_impl = <_abc._abc_data object>
abstract inverse_transform(fixed_image_path: Any, moving_image_path: Any, transformed_image_path: Any, matrix_path: Any, log_file_path: str, interpolator: str = 'default')

Abstract method for inverse transforming images.

Parameters:

  • fixed_image_path (Any) – The fixed image to be transformed.

  • moving_image_path (Any) – The moving image to be transformed.

  • transformed_image_path (Any) – The resulting transformed image.

  • matrix_path (Any) – The transformation matrix applied during transformation.

  • log_file_path (str) – The path to the log file for recording transformation details.

  • interpolator (str) – The interpolator to be used during transformation.

abstract register(fixed_image_path: Any, moving_image_path: Any, transformed_image_path: Any, matrix_path: Any, log_file_path: str)

Abstract method for registering images.

Parameters:

  • fixed_image_path (Any) – The fixed image for registration.

  • moving_image_path (Any) – The moving image to be registered.

  • transformed_image_path (Any) – The resulting transformed image after registration.

  • matrix_path (Any) – The transformation matrix applied during registration.

  • log_file_path (str) – The path to the log file for recording registration details.

abstract transform(fixed_image_path: Any, moving_image_path: Any, transformed_image_path: Any, matrix_path: Any, log_file_path: str, interpolator: str = 'default', **kwargs)

Abstract method for transforming images.

Parameters:

  • fixed_image_path (Any) – The fixed image to be transformed.

  • moving_image_path (Any) – The moving image to be transformed.

  • transformed_image_path (Any) – The resulting transformed image.

  • matrix_path (Any) – The transformation matrix applied during transformation.

  • log_file_path (str) – The path to the log file for recording transformation details.

  • interpolator (str) – The interpolator to be used during transformation.

registration.ANTs

class brainles_preprocessing.registration.ANTs.ANTs.ANTsRegistrator(registration_params: Dict[str, Any] | None = None, transformation_params: Dict[str, Any] | None = None)

Bases: Registrator

_abc_impl = <_abc._abc_data object>
static _log_to_file(log_file_path: str | Path, fixed_image_path: str | Path, moving_image_path: str | Path, transformed_image_path: str | Path, matrix_path: str | Path, operation_name: str, start_time: datetime, end_time: datetime, interpolator: str | None = None) → None

Log the operation details to a file.

Parameters:

  • log_file_path (str or Path) – Path to the log file.

  • fixed_image_path (str or Path) – Path to the fixed image.

  • moving_image_path (str or Path) – Path to the moving image.

  • transformed_image_path (str or Path) – Path to the transformed image.

  • matrix_path (str or Path) – Path to the transformation matrix.

  • operation_name (str) – Name of the operation (‘registration’ or ‘transformation’).

  • start_time (datetime.datetime) – Start time of the operation.

  • end_time (datetime.datetime) – End time of the operation.

  • interpolator (Optional[str]) – Interpolator used for the transformation, if applicable.

inverse_transform(fixed_image_path: str | Path, moving_image_path: str | Path, transformed_image_path: str | Path, matrix_path: str | Path | List[str | Path], log_file_path: str | Path, interpolator: str = 'nearestNeighbor', **kwargs) → None

Apply an inverse transformation using ANTs.

Parameters:

  • fixed_image_path (str or Path) – Path to the fixed image.

  • moving_image_path (str or Path) – Path to the moving image.

  • transformed_image_path (str or Path) – Path to the transformed image (output).

  • matrix_path (str or Path or List[str | Path]) – Path to the transformation matrix or a list of matrices.

  • log_file_path (str or Path) – Path to the log file.

  • interpolator (str) – Interpolator to use for the transformation. Default is ‘nearestNeighbor’.

  • **kwargs – Additional transformation parameters to update the instantiated defaults.

register(fixed_image_path: str | Path, moving_image_path: str | Path, transformed_image_path: str | Path, matrix_path: str | Path, log_file_path: str | Path, **kwargs) → None

Register images using ANTs.

Parameters:

  • fixed_image_path (str or Path) – Path to the fixed image.

  • moving_image_path (str or Path) – Path to the moving image.

  • transformed_image_path (str or Path) – Path to the transformed image (output).

  • matrix_path (str or Path) – Path to the transformation matrix (output).

  • log_file_path (str or Path) – Path to the log file.

  • **kwargs – Additional registration parameters to update the instantiated defaults.

Raises:

FileNotFoundError – If the fixed or moving image paths do not exist.

transform(fixed_image_path: str | Path, moving_image_path: str | Path, transformed_image_path: str | Path, matrix_path: str | Path | List[str | Path], log_file_path: str | Path, interpolator: str = 'nearestNeighbor', **kwargs) → None

Apply a transformation using ANTs. By default the padding value corresponds to the minimum of the moving image. Can be adjusted using the defaultvalue parameter.

Parameters:

  • fixed_image_path (str or Path) – Path to the fixed image.

  • moving_image_path (str or Path) – Path to the moving image.

  • transformed_image_path (str or Path) – Path to the transformed image (output).

  • matrix_path (str or Path or List[str | Path]) – Path to the transformation matrix or a list of matrices.

  • log_file_path (str or Path) – Path to the log file.

  • interpolator (str) – Interpolator to use for the transformation. Default is ‘nearestNeighbor’.

  • **kwargs – Additional transformation parameters to update the instantiated defaults.

Raises:

  • AssertionError – If the interpolator is not valid.

  • FileNotFoundError – If the fixed or moving image paths do not exist.

registration.niftyreg

class brainles_preprocessing.registration.niftyreg.niftyreg.NiftyRegRegistrator(registration_abspath: str = '/home/docs/checkouts/readthedocs.org/user_builds/brainles-preprocessing/checkouts/stable/brainles_preprocessing/registration/niftyreg', registration_script: str | None = None, transformation_script: str | None = None)

Bases: Registrator

_abc_impl = <_abc._abc_data object>
_compose_affine_transforms(transform_paths: List[str | Path], output_path: str | Path) → None

Compose a list of affine transform matrices (4x4), applied in order. i.e., output = Tn @ Tn-1 @ … @ T1

Parameters:

  • transform_paths (list of str or Path) – Paths to .txt files with 4x4 affine matrices.

  • output_path (str or Path) – Where to save the composed transform.

Returns:

None

_invert_affine_transform(transform_path: str | Path, output_path: str | Path) → None

Invert a single affine transform matrix (4x4) and save it.

Parameters:

  • transform_path (str or Path) – Path to the .txt file with the 4x4 affine matrix.

  • output_path (str or Path) – Where to save the inverted transform.

Returns:

None

inverse_transform(fixed_image_path: str, moving_image_path: str, transformed_image_path: str, matrix_path: str | Path | List[str | Path], log_file_path: str, interpolator: str = '0') → None

Apply inverse transformation using NiftyReg.

Parameters:

  • fixed_image_path (str) – Path to the fixed image.

  • moving_image_path (str) – Path to the moving image.

  • transformed_image_path (str) – Path to the transformed image (output).

  • matrix_path (str | Path | List[str | Path]) – Path(s) to the transformation matrix(es) in inverse order.

  • log_file_path (str) – Path to the log file.

  • interpolator (str) – Interpolation order (0, 1, 3, 4) (0=NN, 1=LIN; 3=CUB, 4=SINC), Default is ‘1’ (linear).

register(fixed_image_path: str, moving_image_path: str, transformed_image_path: str, matrix_path: str | Path, log_file_path: str) → None

Register images using NiftyReg.

Parameters:

  • fixed_image_path (str) – Path to the fixed image.

  • moving_image_path (str) – Path to the moving image.

  • transformed_image_path (str) – Path to the transformed image (output).

  • matrix_path (str) – Path to the transformation matrix (output).

  • log_file_path (str) – Path to the log file.

transform(fixed_image_path: str, moving_image_path: str, transformed_image_path: str, matrix_path: str | Path | List[str | Path], log_file_path: str, interpolator: str = '0', **kwargs: dict) → None

Apply a transformation using NiftyReg. By default the padding value corresponds to the minimum of the moving image.

Parameters:

  • fixed_image_path (str) – Path to the fixed image.

  • moving_image_path (str) – Path to the moving image.

  • transformed_image_path (str) – Path to the transformed image (output).

  • matrix_path (str | Path | List[str | Path]) – Path(s) to the transformation matrix(es).

  • log_file_path (str) – Path to the log file.

  • interpolator (str) – Interpolation order (0, 1, 3, 4) (0=NN, 1=LIN; 3=CUB, 4=SINC). Default is ‘1’ (linear).

Raises:

AssertionError – If the interpolator is not valid.

registration.greedy

class brainles_preprocessing.registration.greedy.greedy.GreedyRegistrator

Bases: Registrator

_abc_impl = <_abc._abc_data object>
inverse_transform(fixed_image_path: str, moving_image_path: str, transformed_image_path: str, matrix_path: str, log_file_path: str | None = None, interpolator: str = 'linear') → None

Abstract method for inverse transforming images.

Parameters:

  • fixed_image_path (Any) – The fixed image to be transformed.

  • moving_image_path (Any) – The moving image to be transformed.

  • transformed_image_path (Any) – The resulting transformed image.

  • matrix_path (Any) – The transformation matrix applied during transformation.

  • log_file_path (str) – The path to the log file for recording transformation details.

  • interpolator (str) – The interpolator to be used during transformation.

register(fixed_image_path: str, moving_image_path: str, transformed_image_path: str, matrix_path: str, log_file_path: str | None = None) → None

Register images using greedy. Ref: https://pypi.org/project/picsl-greedy/ and https://greedy.readthedocs.io/en/latest/reference.html#greedy-usage.

This is an optional dependency - to use this registrator, you need to install the brainles_preprocessing package with the picsl_greedy extra: pip install brainles_preprocessing[picsl_greedy].

Parameters:

  • fixed_image_path (str) – Path to the fixed image.

  • moving_image_path (str) – Path to the moving image.

  • transformed_image_path (str) – Path to the transformed image (output).

  • matrix_path (str) – Path to the transformation matrix (output). This gets overwritten if it already exists.

  • log_file_path (Optional[str]) – Path to the log file, which is not used.

transform(fixed_image_path: str, moving_image_path: str, transformed_image_path: str, matrix_path: str, log_file_path: str | None = None, interpolator: str = 'LINEAR', **kwargs: dict | None) → None

Apply a transformation using greedy.

Parameters:

  • fixed_image_path (str) – Path to the fixed image.

  • moving_image_path (str) – Path to the moving image.

  • transformed_image_path (str) – Path to the transformed image (output).

  • matrix_path (str) – Path to the transformation matrix (output). This gets overwritten if it already exists.

  • log_file_path (Optional[str]) – Path to the log file, which is not used.

  • interpolator (Optional[str]) – The interpolator to use; one of NN, LINEAR or LABEL. Defaults to LINEAR.

registration.elastix

class brainles_preprocessing.registration.elastix.elastix.ElastixRegistrator

Bases: Registrator

__initialize_parameter_object() → elastixParameterObject

Initialize the parameter object for elastix registration.

Returns:

The parameter object for registration.

Return type:

itk.ParameterObject

_abc_impl = <_abc._abc_data object>
inverse_transform(fixed_image_path: str, moving_image_path: str, transformed_image_path: str, matrix_path: str, log_file_path: str | None = None, interpolator: str = 'linear') → None

Abstract method for inverse transforming images.

Parameters:

  • fixed_image_path (Any) – The fixed image to be transformed.

  • moving_image_path (Any) – The moving image to be transformed.

  • transformed_image_path (Any) – The resulting transformed image.

  • matrix_path (Any) – The transformation matrix applied during transformation.

  • log_file_path (str) – The path to the log file for recording transformation details.

  • interpolator (str) – The interpolator to be used during transformation.

register(fixed_image_path: str, moving_image_path: str, transformed_image_path: str, matrix_path: str, log_file_path: str | None = None, parameter_object: elastixParameterObject | None = None) → None

Register images using elastix.

This is an optional dependency - to use this registrator, you need to install the brainles_preprocessing package with the itk-elastix extra: pip install brainles_preprocessing[itk-elastix].

Parameters:

  • fixed_image_path (str) – Path to the fixed image.

  • moving_image_path (str) – Path to the moving image.

  • transformed_image_path (str) – Path to the transformed image (output).

  • matrix_path (str) – Path to the transformation matrix (output). This gets overwritten if it already exists.

  • log_file_path (Optional[str]) – Path to the log file.

  • parameter_object (Optional[itk.ParameterObject]) – The parameter object for elastix registration.

transform(fixed_image_path: str, moving_image_path: str, transformed_image_path: str, matrix_path: str, log_file_path: str | None = None, interpolator: str = 'linear', **kwargs) → None

Apply a transformation using elastix.

Parameters:

  • fixed_image_path (str) – Path to the fixed image.

  • moving_image_path (str) – Path to the moving image.

  • transformed_image_path (str) – Path to the transformed image (output).

  • matrix_path (str) – Path to the transformation matrix (output). This gets overwritten if it already exists.

  • log_file_path (Optional[str]) – Path to the log file.