direct.ssl package

direct.ssl package#

Submodules#

direct.ssl.mask_fillers module#

SSL Mask Fillers.

This module contains functions for splitting binary masks into (disjoint) subsets to be used for self-supervised learning MRI reconstruction tasks.

direct.ssl.mask_fillers.gaussian_fill(nonzero_mask_count, nrow, ncol, center_x, center_y, std_scale, mask, output_mask, seed)[source]#

Generates a binary mask filled with randomly sampled positions following a 2D Gaussian distribution.

Makes a call to the cython function _gaussian_fill.

Parameters:

  • nonzero_mask_count (int) – Number of non-zero entries in the output mask.

  • nrow (int) – Number of rows of the output mask.

  • ncol (int) – Number of columns of the output mask.

  • center_x (int) – X coordinate of the center of the Gaussian distribution.

  • center_y (int) – Y coordinate of the center of the Gaussian distribution.

  • std_scale (float) – Scaling factor for the standard deviation of the Gaussian distribution. The standard deviation of the Gaussian distribution will be (nrow-1)/std_scale and (ncol-1)/std_scale along the X and Y axes, respectively.

  • mask (ndarray) – A binary integer 2D array representing the input mask.

  • output_mask (ndarray) – A binary integer 2D array representing the output mask.

  • seed (int) – Seed for the random number generator.

Return type:

ndarray

Returns:

A 2D array representing the output mask filled with randomly sampled positions following a 2D Gaussian distribution.

direct.ssl.mask_fillers.uniform_fill(nonzero_mask_count, nrow, ncol, mask, rng)[source]#

Fills a binary torch.Tensor mask with the specified number of ones in a uniform random manner.

Parameters:

  • nonzero_mask_count (int) – The number of 1s to place in the mask.

  • nrow (int) – The number of rows in the mask.

  • ncol (int) – The number of columns in the mask.

  • mask (Tensor) – A binary mask with zeros and ones.

  • rng (RandomState) – A NumPy random state object for reproducibility.

Return type:

Tensor

Returns:

A binary mask with the specified number of 1s placed in a uniform random manner.

direct.ssl.ssl module#

Direct SSL mask splitters module.

This module contains classes for splitting masks for self-supervised learning tasks for MRI reconstruction. For example, the GaussianMaskSplitterModule splits the input mask into two disjoint masks using a Gaussian split scheme, while the UniformMaskSplitterModule splits the input mask into two disjoint masks using a uniform split scheme. The HalfMaskSplitterModule splits the input mask into two disjoint masks in a half line direction.

class direct.ssl.ssl.GaussianMaskSplitterModule(ratio=0.5, acs_region=(0, 0), keep_acs=False, use_seed=True, kspace_key=KspaceKey.MASKED_KSPACE, std_scale=3.0)[source]#

Bases: MaskSplitter

Uses Gaussian splitting method to split the input mask into two disjoint masks.

Parameters:

  • ratio (Union[float, list[float], tuple[float, ...]]) – Split ratio such that: math:ratio approx frac{|A|}{|B|}. Default: 0.5.

  • acs_region (Union[list[int], tuple[int, int]]) – Size of ACS region to include in training (input) mask. Default: (0, 0).

  • keep_acs (bool) – If True, both input and target masks will keep the acs region and ratio will be applied on the rest of the mask.

  • Default (the same mask every time.) – False.

  • use_seed (bool) – If True, a pseudo-random number based on the filename is computed so that every slice of the volume get

  • Default – True.

  • kspace_key (KspaceKey) – K-space key. Default “masked_kspace”.

  • std_scale (float) – This is used to calculate the standard deviation of the Gaussian distribution. Default: 3.0.

__init__(ratio=0.5, acs_region=(0, 0), keep_acs=False, use_seed=True, kspace_key=KspaceKey.MASKED_KSPACE, std_scale=3.0)[source]#

Inits GaussianMaskSplitterModule.

Parameters:

  • ratio (Union[float, list[float], tuple[float, ...]]) – Split ratio such that: math:ratio approx frac{|A|}{|B|}. Default: 0.5.

  • acs_region (Union[list[int], tuple[int, int]]) – Size of ACS region to include in training (input) mask. Default: (0, 0).

  • keep_acs (bool) – If True, both input and target masks will keep the acs region and ratio will be applied on the rest of the

  • Default (the same mask every time.) – False.

  • use_seed (bool) – If True, a pseudo-random number based on the filename is computed so that every slice of the volume get

  • Default – True.

  • kspace_key (KspaceKey) – K-space key. Default “masked_kspace”.

  • std_scale (float) – This is used to calculate the standard deviation of the Gaussian distribution. Default: 3.0.

split_method(sampling_mask, acs_mask, seed)[source]#

Splits the sampling_mask into two disjoint masks based on gaussian split method.

Parameters:

  • sampling_mask (Tensor) – The input mask tensor to be split.

  • acs_mask (Optional[Tensor]) – The ACS mask. Needs to be passed if keep_acs is True. If keep_acs is False but this is passed, it will

  • Default (be ignored.) – None.

  • seed (Union[int, Iterable[int], None]) – Seed to generate split. Default: None.

Return type:

tuple[Tensor, Tensor]

Returns:

The two disjoint masks, input_mask and target_mask.

class direct.ssl.ssl.HalfMaskSplitterModule(acs_region=(0, 0), keep_acs=False, use_seed=True, direction=HalfSplitType.VERTICAL, kspace_key=KspaceKey.MASKED_KSPACE)[source]#

Bases: MaskSplitter

Splits the input mask into two disjoint masks in a half line direction.

Parameters:

  • acs_region (Union[list[int], tuple[int, int]]) – Size of ACS region to include in training (input) mask. Default: (0, 0).

  • keep_acs (bool) – If True, both input and target masks will keep the acs region and ratio will be applied on the rest of the mask.

  • Default (the same mask every time.) – False.

  • use_seed (bool) – If True, a pseudo-random number based on the filename is computed so that every slice of the volume get

  • Default – True.

  • kspace_key (KspaceKey) – K-space key. Default “masked_kspace”.

__init__(acs_region=(0, 0), keep_acs=False, use_seed=True, direction=HalfSplitType.VERTICAL, kspace_key=KspaceKey.MASKED_KSPACE)[source]#

Inits HalfMaskSplitterModule.

Parameters:

  • acs_region (Union[list[int], tuple[int, int]]) – Size of ACS region to include in training (input) mask. Default: (0, 0).

  • keep_acs (bool) – If True, both input and target masks will keep the acs region and ratio will be applied on the rest of the

  • Default (the same mask every time.) – False.

  • use_seed (bool) – If True, a pseudo-random number based on the filename is computed so that every slice of the volume get

  • Default – True.

  • direction (HalfSplitType) – Direction of the half line split. Default: HalfSplitType.VERTICAL.

  • kspace_key (KspaceKey) – K-space key. Default “masked_kspace”.

split_method(sampling_mask, acs_mask, seed)[source]#

Splits the sampling_mask into two disjoint masks based on half split method.

Parameters:

  • sampling_mask (Tensor) – The input mask tensor to be split.

  • acs_mask (Optional[Tensor]) – The ACS mask. Needs to be passed if keep_acs is True. If keep_acs is False but this is passed, it will

  • Default (be ignored.) – None.

  • seed (Union[int, Iterable[int], None]) – Seed to generate split. Default: None.

Return type:

tuple[Tensor, Tensor]

Returns:

The two disjoint masks, input_mask and target_mask.

class direct.ssl.ssl.HalfSplitType(value, names=_not_given, *values, module=None, qualname=None, type=None, start=1, boundary=None)[source]#

Bases: DirectEnum

SSL half mask splitter types.

These are used to define the type of half mask splitting.

HORIZONTAL = 'horizontal'#
VERTICAL = 'vertical'#
DIAGONAL_LEFT = 'diagonal_left'#
DIAGONAL_RIGHT = 'diagonal_right'#
class direct.ssl.ssl.MaskSplitterType(value, names=_not_given, *values, module=None, qualname=None, type=None, start=1, boundary=None)[source]#

Bases: DirectEnum

SSL mask splitter types.

These are used to define the type of mask splitting.

UNIFORM = 'uniform'#
GAUSSIAN = 'gaussian'#
HALF = 'half'#
class direct.ssl.ssl.UniformMaskSplitterModule(ratio=0.5, acs_region=(0, 0), keep_acs=False, use_seed=True, kspace_key=KspaceKey.MASKED_KSPACE)[source]#

Bases: MaskSplitter

Uses Uniform splitting method to split the input mask into two disjoint masks.

Parameters:

  • ratio (Union[float, list[float], tuple[float, ...]]) – Split ratio such that: math:ratio approx frac{|A|}{|B|}. Default: 0.5.

  • acs_region (Union[list[int], tuple[int, int]]) – Size of ACS region to include in training (input) mask. Default: (0, 0).

  • keep_acs (bool) – If True, both input and target masks will keep the acs region and ratio will be applied on the rest of the mask.

  • Default (the same mask every time.) – False.

  • use_seed (bool) – If True, a pseudo-random number based on the filename is computed so that every slice of the volume get

  • Default – True.

  • kspace_key (KspaceKey) – K-space key. Default “masked_kspace”.

__init__(ratio=0.5, acs_region=(0, 0), keep_acs=False, use_seed=True, kspace_key=KspaceKey.MASKED_KSPACE)[source]#

Inits UniformMaskSplitterModule.

Parameters:

  • ratio (Union[float, list[float], tuple[float, ...]]) – Split ratio such that: math:ratio approx frac{|A|}{|B|}. Default: 0.5.

  • acs_region (Union[list[int], tuple[int, int]]) – Size of ACS region to include in training (input) mask. Default: (0, 0).

  • keep_acs (bool) – If True, both input and target masks will keep the acs region and ratio will be applied on the rest of the

  • Default (the same mask every time.) – False.

  • use_seed (bool) – If True, a pseudo-random number based on the filename is computed so that every slice of the volume get

  • Default – True.

  • kspace_key (KspaceKey) – K-space key. Default “masked_kspace”.

split_method(sampling_mask, acs_mask, seed)[source]#

Splits the sampling_mask into two disjoint masks based on the uniform split method.

Parameters:

  • sampling_mask (Tensor) – The input mask tensor to be split.

  • acs_mask (Optional[Tensor]) – The ACS mask. Needs to be passed if keep_acs is True. If keep_acs is False but this is passed, it will

  • Default (be ignored.) – None.

  • seed (Union[int, Iterable[int], None]) – Seed to generate split. Default: None.

Return type:

tuple[Tensor, Tensor]

Returns:

The two disjoint masks, input_mask and target_mask.

class direct.ssl.ssl.SSLTransformMaskPrefixes(value, names=_not_given, *values, module=None, qualname=None, type=None, start=1, boundary=None)[source]#

Bases: DirectEnum

SSL Transform mask prefixes.

These are used to prefix the input (\(\Theta) and target (:math:\)Lambda`) k-spaces/masks in the sample.

INPUT_ = 'input_'#
TARGET_ = 'target_'#

Module contents#

Direct SSL module.

Includes functions for self-supervised learning for MRI reconstruction.

Contents