def spatial_transformation_volumetric_augmented(self, image):
"""
The spatial image transformation with random augmentation.
:return: The transformation.
"""
dim = 3
transformations_list = [translation.InputCenterToOrigin(dim),
scale.FitFixedAr(dim, self.image_size + [None], ignore_dim=[2])]
if self.scale_factor[0] == 1.0 and self.scale_factor[1] == 1.0:
# if no scale_factor, randomly shift by certain value
transformations_list.append(translation.Random(dim, [self.image_size[0] * 0.35, self.image_size[1] * 0.35, 0.0]))
else:
# else, randomly move in imag size
move_factor = [(1.0 - self.scale_factor[i]) * 0.5 for i in [0, 1]]
transformations_list.append(translation.Random(dim, [self.image_size[0] * move_factor[0], self.image_size[1] * move_factor[1], 0]))
transformations_list.append(scale.Fixed(dim, self.scale_factor + [1.0]))
transformations_list.extend([flip.Random(dim, [0.5, 0.5, 0.0]),
rotation.Random(dim, [0., 0., math.pi]),
scale.RandomUniform(dim, 0.25, ignore_dim=[2]),
scale.Random(dim, [0.25, 0.25, 0.0]),
translation.OriginToOutputCenter(dim, self.image_size + [self.num_frames]),
deformation.Output(dim, [6, 6, 4], [10, 10, 0], self.image_size + [self.num_frames])])
comp = composite.Composite(dim, transformations_list,
name='image_transformation_comp',
kwparents={'image': image})
return LambdaNode(lambda comp: sitk.DisplacementFieldTransform(sitk.TransformToDisplacementField(comp, sitk.sitkVectorFloat64, size=self.image_size + [self.num_frames])),
name='image_transformation',
kwparents={'comp': comp})
def spatial_transformation_augmented(self, image):
"""
The spatial image transformation with random augmentation.
:return: The transformation.
"""
# bring image to center and fit to AR
transformations_list = [translation.InputCenterToOrigin(self.dim),
scale.FitFixedAr(self.dim, self.image_size)]
if self.scale_factor[0] == 1.0 and self.scale_factor[1] == 1.0:
# if no scale_factor, randomly shift by certain value
transformations_list.append(translation.Random(self.dim, [self.image_size[0] * 0.35, self.image_size[1] * 0.35]))
else:
# else, randomly move in imag size
move_factor = [(1.0 - self.scale_factor[i]) * 0.5 for i in [0, 1]]
transformations_list.append(translation.Random(self.dim, [self.image_size[0] * move_factor[0], self.image_size[1] * move_factor[1]]))
transformations_list.append(scale.Fixed(self.dim, self.scale_factor))
transformations_list.extend([flip.Random(self.dim, [0.5, 0.5]),
rotation.Random(self.dim, [math.pi]),
scale.RandomUniform(self.dim, 0.25),
scale.Random(self.dim, [0.25, 0.25]),
translation.OriginToOutputCenter(self.dim, self.image_size),
deformation.Output(self.dim, [8, 8], 10, self.image_size)])
comp = composite.Composite(self.dim, transformations_list,
name='image_transformation_comp',
kwparents={'image': image})
return LambdaNode(lambda comp: sitk.DisplacementFieldTransform(sitk.TransformToDisplacementField(comp, sitk.sitkVectorFloat64, size=self.image_size)),
name='image_transformation',
kwparents={'comp': comp})
def spatial_transformation_augmented(self, iterator, datasources, image_size):
"""
The spatial image transformation with random augmentation.
:param datasources: datasources dict.
:return: The transformation.
"""
transformation_list = []
kwparents = {'image': datasources['image'], 'output_size': image_size}
if self.translate_to_center_landmarks:
kwparents['start'] = datasources['landmarks_bb_start']
kwparents['extent'] = datasources['landmarks_bb_extent']
transformation_list.append(translation.BoundingBoxCenterToOrigin(self.dim, None, self.image_spacing))
elif self.generate_single_vertebrae or self.generate_single_vertebrae_heatmap:
single_landmark = LambdaNode(lambda id_dict, landmarks: [landmarks[int(id_dict['landmark_id'])]],
parents=[iterator, datasources['landmarks']])
kwparents['landmarks'] = single_landmark
transformation_list.append(landmark.Center(self.dim, True))
transformation_list.append(translation.Fixed(self.dim, [0, 20, 0]))
else:
transformation_list.append(translation.InputCenterToOrigin(self.dim))
if self.translate_by_random_factor:
transformation_list.append(translation.RandomCropBoundingBox(self.dim, None, self.image_spacing))
# transformation_list.append(translation.RandomFactorInput(self.dim, [0, 0, 0.5], [0, 0, self.image_spacing[2] * image_size[2]]))
transformation_list.extend([translation.Random(self.dim, [self.random_translation] * self.dim),
rotation.Random(self.dim, [self.random_rotate] * self.dim),
scale.RandomUniform(self.dim, self.random_scale),
scale.Random(self.dim, [self.random_scale] * self.dim),
flip.Random(self.dim, [0.5 if self.random_flip else 0.0, 0.0, 0.0]),
translation.OriginToOutputCenter(self.dim, None, self.image_spacing),
deformation.Output(self.dim, [6, 6, 6], [self.random_deformation] * self.dim, None, self.image_spacing)
])
comp = composite.Composite(self.dim, transformation_list, name='image', kwparents=kwparents)
return LambdaNode(lambda comp, output_size: sitk.DisplacementFieldTransform(sitk.TransformToDisplacementField(comp, sitk.sitkVectorFloat64, size=output_size, outputSpacing=self.image_spacing)),
name='image',
kwparents={'comp': comp, 'output_size': image_size})
def spatial_transformation_augmented(self):
"""
The spatial image transformation with random augmentation.
:return: The transformation.
"""
return composite.Composite(self.dim, [
landmark.Center(self.dim, False),
translation.Random(self.dim, [20, 20, 20]),
rotation.Random(self.dim, [0.35, 0.35, 0.35]),
scale.RandomUniform(self.dim, 0.2),
scale.Random(self.dim, [0.1, 0.1, 0.1]),
translation.OriginToOutputCenter(self.dim, self.image_size,
self.image_spacing),
deformation.Output(self.dim, [8, 8, 8], 15, self.image_size,
self.image_spacing)
])
def spatial_transformation_augmented(self, data_sources):
"""
The spatial image transformation with random augmentation.
:param data_sources: The data_sources dictionary.
:return: The transformation.
"""
return composite.Composite(
self.dim, [
translation.InputCenterToOrigin(self.dim),
translation.Random(self.dim, [10, 10]),
rotation.Random(self.dim, [0.25]),
scale.RandomUniform(self.dim, 0.2),
scale.Random(self.dim, [0.2, 0.2]),
translation.OriginToOutputCenter(self.dim, self.image_size,
self.image_spacing),
deformation.Output(self.dim, [5, 5], 10, self.image_size,
self.image_spacing)
],
kwparents={'image': data_sources['image_datasource']},
name='image')
def spatial_transformation_augmented(self, datasources):
"""
The spatial image transformation with random augmentation.
:param datasources: datasources dict.
:return: The transformation.
"""
transformation_list = []
kwparents = {'image': datasources['image']}
transformation_list.append(translation.InputCenterToOrigin(self.dim))
transformation_list.extend([
translation.Random(self.dim, [20, 20, 20]),
rotation.Random(self.dim, [0.35, 0.35, 0.35]),
scale.RandomUniform(self.dim, 0.2),
scale.Random(self.dim, [0.1, 0.1, 0.1]),
translation.OriginToOutputCenter(self.dim, self.image_size,
self.image_spacing),
deformation.Output(self.dim, [6, 6, 6], 15, self.image_size,
self.image_spacing)
])
return composite.Composite(self.dim,
transformation_list,
name='image',
kwparents=kwparents)