def test_tags_to_anonymise_in_dicom_dict_baseline(
save_new_identifying_keywords=False, save_new_baselines=False):
baseline_keywords = [val[4] for val in get_baseline_dicom_dict().values()]
assert set(IDENTIFYING_KEYWORDS).issubset(baseline_keywords)
if save_new_identifying_keywords:
with open(IDENTIFYING_KEYWORDS_FILEPATH, "w") as outfile:
json.dump(IDENTIFYING_KEYWORDS, outfile, indent=2, sort_keys=True)
# TODO: Keywords to add if/when anonymisation of UIDs is implemented:
# "AffectedSOPInstanceUID",
# "ConcatenationUID",
# "ContextGroupExtensionCreatorUID",
# "CreatorVersionUID",
# "DeviceUID",
# "DigitalSignatureUID",
# "DimensionOrganizationUID",
# "DoseReferenceUID",
# "FailedSOPInstanceUIDList",
# "FiducialUID",
# "FrameOfReferenceUID",
# "InstanceCreatorUID",
# "IrradiationEventUID",
# "LargePaletteColorLookupTableUID",
# "MediaStorageSOPInstanceUID",
# "PaletteColorLookupTableUID",
# "ReferencedFrameOfReferenceUID",
# "ReferencedGeneralPurposeScheduledProcedureStepTransactionUID",
# "ReferencedSOPInstanceUID",
# "ReferencedSOPInstanceUIDInFile",
# "RelatedFrameOfReferenceUID",
# "RequestedSOPInstanceUID",
# "SeriesInstanceUID",
# "SOPInstanceUID",
# "StorageMediaFileSetUID",
# "StudyInstanceUID",
# "SynchronizationFrameOfReferenceUID",
# "TemplateExtensionCreatorUID",
# "TemplateExtensionOrganizationUID",
# "TransactionUID",
# "UID",
if save_new_baselines:
with open(BASELINE_DICOM_DICT_FILEPATH, "w") as outfile:
json.dump(get_baseline_dicom_dict(),
outfile,
indent=2,
sort_keys=True)
with open(BASELINE_DICOM_REPEATERS_DICT_FILEPATH, "w") as outfile:
json.dump(get_baseline_dicom_repeaters_dict(),
outfile,
indent=2,
sort_keys=True)
def test_tags_to_anonymise_in_dicom_dict_baseline(save_new_identifying_keywords=False):
baseline_keywords = [val[4] for val in get_baseline_dicom_dict().values()]
assert set(IDENTIFYING_KEYWORDS).issubset(baseline_keywords)
if save_new_identifying_keywords:
with open(IDENTIFYING_KEYWORDS_FILEPATH, "w") as outfile:
json.dump(IDENTIFYING_KEYWORDS, outfile, indent=2, sort_keys=True)
def test_anonymise_dataset_and_all_is_anonymised_functions(tmp_path):
# Create dataset with one instance of every identifying keyword and
# run basic anonymisation tests
ds = pydicom.dataset.Dataset()
for keyword in IDENTIFYING_KEYWORDS:
# Ignore file meta elements for now
tag = hex(pydicom.datadict.tag_for_keyword(keyword))
if pydicom.tag.Tag(tag).group == 0x0002:
continue
value = _get_non_anonymous_replacement_value(keyword)
setattr(ds, keyword, value)
_check_is_anonymised_dataset_file_and_dir(ds, tmp_path, anon_is_expected=False)
ds_anon = anonymise_dataset(ds)
_check_is_anonymised_dataset_file_and_dir(ds_anon, tmp_path, anon_is_expected=True)
# Test the anonymisation and check functions for each identifying
# element individually.
for elem in ds_anon.iterall():
# TODO: AffectedSOPInstanceUID and RequestedSOPInstanceUID
# are not writing to file. Investigate when UID anonymisation is
# implemented.
if elem.keyword in ("AffectedSOPInstanceUID", "RequestedSOPInstanceUID"):
continue
ds_single_non_anon_value = deepcopy(ds_anon)
setattr(
ds_single_non_anon_value,
elem.keyword,
_get_non_anonymous_replacement_value(elem.keyword),
)
_check_is_anonymised_dataset_file_and_dir(
ds_single_non_anon_value, tmp_path, anon_is_expected=False
)
ds_single_anon = anonymise_dataset(ds_single_non_anon_value)
_check_is_anonymised_dataset_file_and_dir(
ds_single_anon, tmp_path, anon_is_expected=True
)
# Test correct handling of private tags
ds_anon.add(pydicom.dataset.DataElement(0x0043102B, "SS", [4, 4, 0, 0]))
_check_is_anonymised_dataset_file_and_dir(
ds_anon, tmp_path, anon_is_expected=False, ignore_private_tags=False
)
_check_is_anonymised_dataset_file_and_dir(
ds_anon, tmp_path, anon_is_expected=True, ignore_private_tags=True
)
ds_anon.remove_private_tags()
_check_is_anonymised_dataset_file_and_dir(
ds_anon, tmp_path, anon_is_expected=True, ignore_private_tags=False
)
# Test blank anonymisation
# # Sanity check
_check_is_anonymised_dataset_file_and_dir(ds, tmp_path, anon_is_expected=False)
ds_anon_blank = anonymise_dataset(ds, replace_values=False)
_check_is_anonymised_dataset_file_and_dir(
ds_anon_blank, tmp_path, anon_is_expected=True
)
# Test handling of unknown tags by removing PatientName from
# baseline dict
patient_name_tag = pydicom.datadict.tag_for_keyword("PatientName")
try:
patient_name = get_baseline_dicom_dict().pop(patient_name_tag)
with pytest.raises(ValueError) as e_info:
anonymise_dataset(ds)
assert str(e_info.value).count(
"At least one of the non-private tags "
"within your DICOM file is not within "
"PyMedPhys's copy of the DICOM dictionary."
)
ds_anon_delete_unknown = anonymise_dataset(ds, delete_unknown_tags=True)
_check_is_anonymised_dataset_file_and_dir(
ds_anon_delete_unknown, tmp_path, anon_is_expected=True
)
with pytest.raises(AttributeError) as e_info:
ds_anon_delete_unknown.PatientName # pylint: disable = pointless-statement
assert str(e_info.value).count(
"'Dataset' object has no attribute " "'PatientName'"
)
ds_anon_ignore_unknown = anonymise_dataset(ds, delete_unknown_tags=False)
_check_is_anonymised_dataset_file_and_dir(
ds_anon_ignore_unknown, tmp_path, anon_is_expected=True
)
assert patient_name_tag in ds_anon_ignore_unknown
finally:
get_baseline_dicom_dict().setdefault(patient_name_tag, patient_name)
# Test copy_dataset=False:
anonymise_dataset(ds, copy_dataset=False)
assert is_anonymised_dataset(ds)
