def _test_Crop(im_seg, crop_transform):
im, seg = im_seg
metadata_ = SampleMetadata({'data_shape': im[0].shape, 'crop_params': {}})
metadata_in = [metadata_ for _ in im] if isinstance(im, list) else {}
if crop_transform.__class__.__name__ == "ROICrop":
_, metadata_in = crop_transform(seg, metadata_in)
for metadata in metadata_in:
assert crop_transform.__class__.__name__ in metadata["crop_params"]
# Apply transform
do_im, do_metadata = crop_transform(im, metadata_in)
do_seg, do_seg_metadata = crop_transform(seg, metadata_in)
crop_transfrom_size = crop_transform.size if not len(
do_im[0].shape) == 2 else crop_transform.size[:2]
# Loop and check
for idx, i in enumerate(im):
# Check data shape
assert list(do_im[idx].shape) == crop_transfrom_size
assert list(do_seg[idx].shape) == crop_transfrom_size
# Check metadata
assert do_metadata[idx]['crop_params'][crop_transform.__class__.__name__] == \
do_seg_metadata[idx]['crop_params'][crop_transform.__class__.__name__]
# Apply undo transform
undo_im, _ = crop_transform.undo_transform(do_im, do_metadata)
undo_seg, _ = crop_transform.undo_transform(do_seg, do_seg_metadata)
# Check data type and shape
_check_dtype(im, [undo_im])
_check_shape(im, [undo_im])
_check_dtype(seg, [undo_seg])
_check_shape(seg, [undo_seg])
# Loop and check
for idx, i in enumerate(im):
# Check data consistency
fh, fw, fd, _, _, _ = do_metadata[idx]['crop_params'][
crop_transform.__class__.__name__]
th, tw, td = crop_transform.size
if not td:
assert np.array_equal(i[fh:fh + th, fw:fw + tw],
undo_im[idx][fh:fh + th, fw:fw + tw])
assert np.array_equal(seg[idx][fh:fh + th, fw:fw + tw],
undo_seg[idx][fh:fh + th, fw:fw + tw])
# Plot for debugging
if DEBUGGING:
plot_transformed_sample(seg[idx], undo_seg[idx],
['raw', 'undo'])
plot_transformed_sample(i, undo_im[idx], ['raw', 'undo'])
else:
assert np.array_equal(
i[fh:fh + th, fw:fw + tw, fd:fd + td],
undo_im[idx][fh:fh + th, fw:fw + tw, fd:fd + td])
assert np.array_equal(
seg[idx][fh:fh + th, fw:fw + tw, fd:fd + td],
undo_seg[idx][fh:fh + th, fw:fw + tw, fd:fd + td])
def test_Clahe(im_seg, clahe):
im, seg = im_seg
metadata_in = [SampleMetadata({}) for _ in im] if isinstance(im, list) else SampleMetadata({})
# Transform on Numpy
do_im, metadata_do = clahe(im.copy(), metadata_in)
_check_dtype(im, [do_im])
_check_shape(im, [do_im])
if DEBUGGING and len(im[0].shape) == 2:
plot_transformed_sample(im[0], do_im[0], ['raw', 'do'])
def _test_Resample(im_seg, resample_transform, native_resolution, is_2D=False):
im, seg = im_seg
metadata_ = SampleMetadata({
'zooms':
native_resolution,
'data_shape':
im[0].shape if len(im[0].shape) == 3 else list(im[0].shape) + [1],
'data_type':
'im'
})
metadata_in = [metadata_
for _ in im] if isinstance(im, list) else SampleMetadata({})
# Resample input data
do_im, do_metadata = resample_transform(sample=im, metadata=metadata_in)
# Undo Resample on input data
undo_im, _ = resample_transform.undo_transform(sample=do_im,
metadata=do_metadata)
# Resampler for label data
resample_transform.interpolation_order = 0
metadata_ = SampleMetadata({
'zooms':
native_resolution,
'data_shape':
seg[0].shape if len(seg[0].shape) == 3 else list(seg[0].shape) + [1],
'data_type':
'gt'
})
metadata_in = [metadata_ for _ in seg] if isinstance(
seg, list) else SampleMetadata({})
# Resample label data
do_seg, do_metadata = resample_transform(sample=seg, metadata=metadata_in)
# Undo Resample on label data
undo_seg, _ = resample_transform.undo_transform(sample=do_seg,
metadata=do_metadata)
# Check data type and shape
_check_dtype(im, [undo_im])
_check_shape(im, [undo_im])
_check_dtype(seg, [undo_seg])
_check_shape(seg, [undo_seg])
# Check data content and data shape between input data and undo
for idx, i in enumerate(im):
# Plot for debugging
if DEBUGGING and is_2D:
plot_transformed_sample(im[idx], undo_im[idx], ['raw', 'undo'])
plot_transformed_sample(seg[idx], undo_seg[idx], ['raw', 'undo'])
# Data consistency
assert dice_score(undo_seg[idx], seg[idx]) > 0.8
def test_RandomAffine(im_seg, transform):
im, seg = im_seg
metadata_in = [SampleMetadata({})
for _ in im] if isinstance(im, list) else SampleMetadata({})
# Transform on Numpy
do_im, metadata_do = transform(im.copy(), metadata_in)
do_seg, metadata_do = transform(seg.copy(), metadata_do)
if DEBUGGING and len(im[0].shape) == 2:
plot_transformed_sample(im[0], do_im[0], ['raw', 'do'])
plot_transformed_sample(seg[0], do_seg[0], ['raw', 'do'])
# Transform on Numpy
undo_im, _ = transform.undo_transform(do_im, metadata_do)
undo_seg, _ = transform.undo_transform(do_seg, metadata_do)
if DEBUGGING and len(im[0].shape) == 2:
# TODO: ERROR for image but not for seg.....
plot_transformed_sample(im[0], undo_im[0], ['raw', 'undo'])
plot_transformed_sample(seg[0], undo_seg[0], ['raw', 'undo'])
# Check data type and shape
_check_dtype(im, [do_im, undo_im])
_check_shape(im, [do_im, undo_im])
_check_dtype(seg, [undo_seg, do_seg])
_check_shape(seg, [undo_seg, do_seg])
# Loop and check
for idx, i in enumerate(im):
# Data consistency
assert dice_score(undo_seg[idx], seg[idx]) > 0.85
def test_ElasticTransform(im_seg, elastic_transform):
im, seg = im_seg
metadata_in = [SampleMetadata({}) for _ in im] if isinstance(im, list) else SampleMetadata({})
# Transform on Numpy
do_im, metadata_do = elastic_transform(im.copy(), metadata_in)
do_seg, metadata_do = elastic_transform(seg.copy(), metadata_do)
if DEBUGGING and len(im[0].shape) == 2:
plot_transformed_sample(im[0], do_im[0], ['raw', 'do'])
plot_transformed_sample(seg[0], do_seg[0], ['raw', 'do'])
_check_dtype(im, [do_im])
_check_shape(im, [do_im])
_check_dtype(seg, [do_seg])
_check_shape(seg, [do_seg])
def test_DilateGT(im_seg, dilate_transform):
im, seg = im_seg
metadata_in = [SampleMetadata({}) for _ in im] if isinstance(im, list) else SampleMetadata({})
# Transform on Numpy
do_seg, metadata_do = dilate_transform(seg.copy(), metadata_in)
if DEBUGGING and len(im[0].shape) == 2:
plot_transformed_sample(seg[0], do_seg[0], ['raw', 'do'])
# Check data shape and type
_check_shape(ref=seg, list_mov=[do_seg])
# Check data augmentation
for idx, i in enumerate(seg):
# data aug
assert np.sum((do_seg[idx] > 0).astype(int)) >= np.sum(i)
# same number of objects
assert label((do_seg[idx] > 0).astype(int))[1] == label(i)[1]
def test_RandomReverse(im_seg, reverse_transform):
im, seg = im_seg
metadata_in = [SampleMetadata({}) for _ in im] if isinstance(im, list) else SampleMetadata({})
# Transform on Numpy
do_im, metadata_do = reverse_transform(im.copy(), metadata_in)
do_seg, metadata_do = reverse_transform(seg.copy(), metadata_do)
# Transform on Numpy
undo_im, _ = reverse_transform.undo_transform(do_im, metadata_do)
undo_seg, _ = reverse_transform.undo_transform(do_seg, metadata_do)
if DEBUGGING and len(im[0].shape) == 2:
plot_transformed_sample(seg[0], do_seg[0], ['raw', 'do'])
plot_transformed_sample(seg[0], undo_seg[0], ['raw', 'undo'])
_check_dtype(im, [do_im])
_check_shape(im, [do_im])
_check_dtype(seg, [do_seg])
_check_shape(seg, [do_seg])
def run_visualization(input, config, number, output, roi):
"""Utility function to visualize Data Augmentation transformations.
Data augmentation is a key part of the Deep Learning training scheme. This script aims at facilitating the
fine-tuning of data augmentation parameters. To do so, this script provides a step-by-step visualization of the
transformations that are applied on data.
This function applies a series of transformations (defined in a configuration file
``-c``) to ``-n`` 2D slices randomly extracted from an input image (``-i``), and save as png the resulting sample
after each transform.
For example::
ivadomed_visualize_transforms -i t2s.nii.gz -n 1 -c config.json -r t2s_seg.nii.gz
Provides a visualization of a series of three transformation on a randomly selected slice:
.. image:: https://raw.githubusercontent.com/ivadomed/doc-figures/main/scripts/transforms_im.png
:width: 600px
:align: center
And on a binary mask::
ivadomed_visualize_transforms -i t2s_gmseg.nii.gz -n 1 -c config.json -r t2s_seg.nii.gz
Gives:
.. image:: https://raw.githubusercontent.com/ivadomed/doc-figures/main/scripts/transforms_gt.png
:width: 600px
:align: center
Args:
input (string): Image filename. Flag: ``--input``, ``-i``
config (string): Configuration file filename. Flag: ``--config``, ``-c``
number (int): Number of slices randomly extracted. Flag: ``--number``, ``-n``
output (string): Folder path where the results are saved. Flag: ``--ofolder``, ``-o``
roi (string): Filename of the region of interest. Only needed if ROICrop is part of the transformations.
Flag: ``--roi``, ``-r``
"""
# Load context
context = imed_config_manager.ConfigurationManager(config).get_config()
# Create output folder
if not Path(output).is_dir():
Path(output).mkdir(parents=True)
# Slice extracted according to below axis
axis = imed_utils.AXIS_DCT[context[ConfigKW.LOADER_PARAMETERS][LoaderParamsKW.SLICE_AXIS]]
# Get data
input_img, input_data = get_data(input, axis)
# Image or Mask
is_mask = np.array_equal(input_data, input_data.astype(bool))
# Get zooms
zooms = imed_loader_utils.orient_shapes_hwd(input_img.header.get_zooms(), slice_axis=axis)
# Get indexes
indexes = random.sample(range(0, input_data.shape[2]), number)
# Get training transforms
training_transforms, _, _ = imed_transforms.get_subdatasets_transforms(context[ConfigKW.TRANSFORMATION])
if TransformationKW.ROICROP in training_transforms:
if roi and Path(roi).is_file():
roi_img, roi_data = get_data(roi, axis)
else:
raise ValueError("\nPlease provide ROI image (-r) in order to apply ROICrop transformation.")
# Compose transforms
dict_transforms = {}
stg_transforms = ""
for transform_name in training_transforms:
# We skip NumpyToTensor transform since that s only a change of data type
if transform_name == "NumpyToTensor":
continue
# Update stg_transforms
stg_transforms += transform_name + "_"
# Add new transform to Compose
dict_transforms.update({transform_name: training_transforms[transform_name]})
composed_transforms = imed_transforms.Compose(dict_transforms)
# Loop across slices
for i in indexes:
data = [input_data[:, :, i]]
# Init metadata
metadata = SampleMetadata({MetadataKW.ZOOMS: zooms, MetadataKW.DATA_TYPE: "gt" if is_mask else "im"})
# Apply transformations to ROI
if TransformationKW.CENTERCROP in training_transforms or \
(TransformationKW.ROICROP in training_transforms and Path(roi).is_file()):
metadata.__setitem__(MetadataKW.CROP_PARAMS, {})
# Apply transformations to image
stack_im, _ = composed_transforms(sample=data,
metadata=[metadata for _ in range(number)],
data_type="im")
# Plot before / after transformation
fname_out = str(Path(output, stg_transforms + "slice" + str(i) + ".png"))
logger.debug(f"Fname out: {fname_out}.")
logger.debug(f"\t{dict(metadata)}")
# rescale intensities
if len(stg_transforms[:-1].split("_")) == 1:
before = np.rot90(imed_maths.rescale_values_array(data[0], 0.0, 1.0))
else:
before = after
if isinstance(stack_im[0], torch.Tensor):
after = np.rot90(imed_maths.rescale_values_array(stack_im[0].numpy(), 0.0, 1.0))
else:
after = np.rot90(imed_maths.rescale_values_array(stack_im[0], 0.0, 1.0))
# Plot
imed_utils.plot_transformed_sample(before,
after,
list_title=["\n".join(stg_transforms[:-1].split("_")[:-1]),
"\n".join(stg_transforms[:-1].split("_"))],
fname_out=fname_out,
cmap="jet" if is_mask else "gray")
def run_visualization(input, config, number, output, roi):
"""Utility function to visualize Data Augmentation transformations.
Data augmentation is a key part of the Deep Learning training scheme. This script aims at facilitating the
fine-tuning of data augmentation parameters. To do so, this script provides a step-by-step visualization of the
transformations that are applied on data.
This function applies a series of transformations (defined in a configuration file
``-c``) to ``-n`` 2D slices randomly extracted from an input image (``-i``), and save as png the resulting sample
after each transform.
For example::
ivadomed_visualize_transforms -i t2s.nii.gz -n 1 -c config.json -r t2s_seg.nii.gz
Provides a visualization of a series of three transformation on a randomly selected slice:
.. image:: ../../images/transforms_im.png
:width: 600px
:align: center
And on a binary mask::
ivadomed_visualize_transforms -i t2s_gmseg.nii.gz -n 1 -c config.json -r t2s_seg.nii.gz
Gives:
.. image:: ../../images/transforms_gt.png
:width: 600px
:align: center
Args:
input (string): Image filename. Flag: ``--input``, ``-i``
config (string): Configuration file filename. Flag: ``--config``, ``-c``
number (int): Number of slices randomly extracted. Flag: ``--number``, ``-n``
output (string): Folder path where the results are saved. Flag: ``--ofolder``, ``-o``
roi (string): Filename of the region of interest. Only needed if ROICrop is part of the transformations.
Flag: ``--roi``, ``-r``
"""
# Load context
with open(config, "r") as fhandle:
context = json.load(fhandle)
# Create output folder
if not os.path.isdir(output):
os.makedirs(output)
# Slice extracted according to below axis
axis = imed_utils.AXIS_DCT[context["loader_parameters"]["slice_axis"]]
# Get data
input_img, input_data = get_data(input, axis)
# Image or Mask
is_mask = np.array_equal(input_data, input_data.astype(bool))
# Get zooms
zooms = imed_loader_utils.orient_shapes_hwd(input_img.header.get_zooms(),
slice_axis=axis)
# Get indexes
indexes = random.sample(range(0, input_data.shape[2]), number)
# Get training transforms
training_transforms, _, _ = imed_transforms.get_subdatasets_transforms(
context["transformation"])
if "ROICrop" in training_transforms:
if roi and os.path.isfile(roi):
roi_img, roi_data = get_data(roi, axis)
else:
print(
"\nPlease provide ROI image (-r) in order to apply ROICrop transformation."
)
exit()
# Compose transforms
dict_transforms = {}
stg_transforms = ""
for transform_name in training_transforms:
# We skip NumpyToTensor transform since that s only a change of data type
if transform_name == "NumpyToTensor":
continue
# Update stg_transforms
stg_transforms += transform_name + "_"
# Add new transform to Compose
dict_transforms.update(
{transform_name: training_transforms[transform_name]})
composed_transforms = imed_transforms.Compose(dict_transforms)
# Loop across slices
for i in indexes:
data = [input_data[:, :, i]]
# Init metadata
metadata = imed_loader_utils.SampleMetadata({
"zooms":
zooms,
"data_type":
"gt" if is_mask else "im"
})
# Apply transformations to ROI
if "CenterCrop" in training_transforms or (
"ROICrop" in training_transforms and os.path.isfile(roi)):
metadata.__setitem__('crop_params', {})
# Apply transformations to image
stack_im, _ = composed_transforms(
sample=data,
metadata=[metadata for _ in range(number)],
data_type="im")
# Plot before / after transformation
fname_out = os.path.join(
output, stg_transforms + "slice" + str(i) + ".png")
print("Fname out: {}.".format(fname_out))
print("\t{}".format(dict(metadata)))
# rescale intensities
if len(stg_transforms[:-1].split("_")) == 1:
before = np.rot90(
imed_maths.rescale_values_array(data[0], 0.0, 1.0))
else:
before = after
after = np.rot90(
imed_maths.rescale_values_array(stack_im[0], 0.0, 1.0))
# Plot
imed_utils.plot_transformed_sample(
before,
after,
list_title=[
"\n".join(stg_transforms[:-1].split("_")[:-1]),
"\n".join(stg_transforms[:-1].split("_"))
],
fname_out=fname_out,
cmap="jet" if is_mask else "gray")
