def default_patient_mask(self):
self.patient_mask.StudyDate = '' # (0008,0020)
self.patient_mask.PatientName = '' # (0010,0010)
self.patient_mask.PatientBirthDate = '' # (0010,0030)
self.patient_mask.PatientID = '' # (0010,0020)
self.patient_mask.StudyDescription = '' # (0008,1030)
self.patient_mask.AccessionNumber = '' # (0008,0050)
self.patient_mask.ReferringPhysicianName = '' # (0008,0090)
self.patient_mask.add(DataElement(0x00081060, 'PN', '')) # (0008,1060)
self.patient_mask.add(DataElement(0x00080080, 'LO',
'')) # InstitutionName
self.patient_mask.add(DataElement(0x00081050, 'PN',
'')) # PerformingPhysicianName
self.patient_mask.StudyInstanceUID = ''
self.patient_mask.QueryRetrieveLevel = 'STUDY'
def create_attribute_list(attrs):
ds = Dataset()
ds.PatientID = '12345'
ds.PatientName = 'Test^User'
for attr in attrs:
print('(VR, VM, Name, Retired, Keyword)\n', DicomDictionary.get(attr)) #
didic = DicomDictionary.get(attr)
elem = DataElement(attr, didic[0], '')
ds.add(elem)
return ds
def convert(self):
# create file meta information
file_meta = Dataset()
file_meta.MediaStorageSOPClassUID = '1.2.840.10008.5.1.4.1.1.77.1.2' # VL Microscopic Image Storage
file_meta.MediaStorageSOPInstanceUID = "1.2.276.0.7230010.3.1.4.296485376.1.1484917438.721089"
file_meta.ImplementationClassUID = "1.2.3.4"
file_meta.FileMetaInformationVersion = b'\x00\x01'
file_meta.FileMetaInformationGroupLength = len(file_meta)
if self.JPEG_COMPRESS:
# file_meta.TransferSyntaxUID = '1.2.840.10008.1.2.4.80' # JPEG 2k
# file_meta.TransferSyntaxUID = '1.2.840.10008.1.2.4.70' # JPEG
file_meta.TransferSyntaxUID = '1.2.840.10008.1.2.4.50' # JPEG baseline
else:
file_meta.TransferSyntaxUID = '1.2.840.10008.1.2' # default uncompressed
# write data into Dicom instances
self.instance_cnt = 0
for frame_items_info in self.frame_items_info_list:
print("Saving to instance %d/%d" %
(self.instance_cnt, len(self.frame_items_info_list)))
# update relevant tags
self.dcm_instance.InstanceNumber = self.instance_cnt
self.dcm_instance.SeriesInstanceUID = '1.2.276.0.7230010.3.1.3.296485376.1.1484917433.721085.' + str(
frame_items_info.img_level)
self.dcm_instance.SeriesNumber = frame_items_info.img_level
print(frame_items_info.img_level)
# self.dcm_instance.SOPInstanceUID = self.dcm_instance.SOPInstanceUID + str(self.instance_cnt)
self.dcm_instance.SOPInstanceUID = '1.2.276.0.7230010.3.1.4.296485376.1.1484917438.721089.' + str(
self.instance_cnt)
self.dcm_instance.NumberOfFrames = len(frame_items_info.locations)
self.dcm_instance.TotalPixelMatrixColumns, self.dcm_instance.TotalPixelMatrixRows = self.wsi_obj.level_dimensions[
frame_items_info.img_level]
self.add_Frame_Sequence_data(frame_items_info)
# create encoded pixel data
PixelData_encoded = self.add_PixelData(frame_items_info)
if self.JPEG_COMPRESS:
filename = os.path.join(
self.save_to_dir,
"compressed_instance_" + str(self.instance_cnt) + ".dcm")
data_elem_tag = pydicom.tag.TupleTag((0x7FE0, 0x0010))
enc_frames = encapsulate(PixelData_encoded, has_bot=True)
pd_ele = DataElement(data_elem_tag,
'OB',
enc_frames,
is_undefined_length=True)
self.dcm_instance.add(pd_ele)
else:
filename = os.path.join(
self.save_to_dir,
"instance_" + str(self.instance_cnt) + ".dcm")
self.dcm_instance.PixelData = PixelData_encoded
self.dcm_instance.file_meta = file_meta
self.dcm_instance.save_as(filename, write_like_original=False)
self.instance_cnt += 1
def add_graphic_annotation(dicom, group_number, layer, type, origin, rows,
columns, data):
dicom[group_number, 0x1001] = DataElement(Tag(group_number, 0x1001), "CS",
layer)
dicom[group_number, 0x40] = DataElement(Tag(group_number, 0x40), "CS",
type)
dicom[group_number, 0x50] = DataElement(Tag(group_number, 0x50), "SS",
origin)
dicom[group_number, 0x10] = DataElement(Tag(group_number, 0x10), "US",
rows)
dicom[group_number, 0x11] = DataElement(Tag(group_number, 0x11), "US",
columns)
dicom[group_number, 0x100] = DataElement(Tag(group_number, 0x100), "US",
1) # OverlayBitsAllocated
dicom[group_number, 0x102] = DataElement(Tag(group_number, 0x102), "US",
0) # OverlayBitPosition
dicom[group_number, 0x3000] = DataElement(Tag(group_number, 0x3000), "OW",
data)
return dicom
def writeDicom(dataset, outfilename):
datasetObj = json.loads(dataset.encode('GB18030').decode('iso8859'))
# 读取模板Dicom文件
filename = get_testdata_files('color-px.dcm')[0]
ds = pydicom.dcmread(filename)
# 当前时间
dt = datetime.datetime.now()
# print("Setting file meta information...")
# Populate required values for file meta information
# file_meta = ds.file_meta
# file_meta.MediaStorageSOPClassUID = '1.2.840.10008.5.1.4.1.1.2'
# file_meta.MediaStorageSOPInstanceUID = "1.2.3"
# file_meta.ImplementationClassUID = "1.2.3.4"
# file_meta.TransferSyntaxUID = "1.2.840.10008.1.2.5"
print("Setting dataset values...")
# Create the FileDataset instance (initially no data elements, but file_meta supplied)
# ds = FileDataset(filename, {}, file_meta=file_meta, preamble=b"\0" * 128)
img = Image.open(datasetObj['ImagePixel']['PixelData'])
nparr = np.asarray(img)
# Pixel Data
# ds[0x7FE0, 0x0010] = DataElement(0x7FE00010, 'OW', img.tobytes())
ds.PixelData = nparr.tobytes()
# Patient's Name
ds[0x0010, 0x0010] = DataElement(0x00100010, 'PN',
datasetObj['Patient']['PatientName'])
# Patient ID
ds[0x0010, 0x0020] = DataElement(0x00100020, 'LO',
datasetObj['Patient']['PatientID'])
# Patient's Birth Date
ds[0x0010, 0x0030] = DataElement(0x00100030, 'DA',
datasetObj['Patient']['PatientBirthDate'])
# Patient's Sex
ds[0x0010, 0x0040] = DataElement(0x00100040, 'CS',
datasetObj['Patient']['PatientSex'])
# Study Date
ds[0x0008, 0x0020] = DataElement(0x00080020, 'DA',
datasetObj['GeneralStudy']['StudyDate'])
# Study Time
ds[0x0008, 0x0030] = DataElement(0x00080030, 'TM',
datasetObj['GeneralStudy']['StudyTime'])
# Accession Number
ds[0x0008,
0x0050] = DataElement(0x00080050, 'SH',
datasetObj['GeneralStudy']['AccessionNumber'])
# Study Instance UID
# ds[0x0020, 0x000D] = DataElement(0x0020000D, 'UI', '1.2.840.1.2.8.236.511020181107')
# Study ID
ds[0x0020, 0x0010] = DataElement(0x00200010, 'SH',
datasetObj['GeneralStudy']['StudyID'])
# Patient's Age
ds[0x0010, 0x1010] = DataElement(0x00101010, 'AS',
datasetObj['PatientStudy']['PatientAge'])
# Patient's Size
ds[0x0010, 0x1020] = DataElement(0x00101020, 'DS',
datasetObj['PatientStudy']['PatientSize'])
# Patient's Weight
ds[0x0010,
0x1030] = DataElement(0x00101030, 'DS',
datasetObj['PatientStudy']['PatientWeight'])
# Series Date
ds[0x0008, 0x0021] = DataElement(0x00080021, 'DA', dt.strftime('%Y%m%d'))
# Series Time
ds[0x0008, 0x0031] = DataElement(0x00080031, 'TM',
dt.strftime('%H%M%S.%f'))
# Modality
ds[0x0008, 0x0060] = DataElement(0x00080060, 'CS',
datasetObj['GeneralSeries']['Modality'])
# Series Description
ds[0x0008, 0x103E] = DataElement(0x0008103E, 'LO', '')
# Performing Physician's Name
ds[0x0008, 0x1050] = DataElement(0x00081050, 'PN', '')
# Body Part Examined
ds[0x0018, 0x0015] = DataElement(0x00180015, 'CS', '')
# Series Instance UID
# ds[0x0020, 0x000E] = DataElement(0x0020000E, 'UI', '1.2.840.1.2.8.236.51102018110715')
# Series Number
ds[0x0020,
0x0011] = DataElement(0x00200011, 'IS',
datasetObj['GeneralSeries']['SeriesNumber'])
# Manufacturer
ds[0x0008,
0x0070] = DataElement(0x00080070, 'LO',
datasetObj['GeneralEquipment']['Manufacturer'])
# Institution Name
ds[0x0008,
0x0080] = DataElement(0x00080080, 'LO',
datasetObj['GeneralEquipment']['InstitutionName'])
# Image Type
# ds[0x0008, 0x0008] = DataElement(0x00080008, 'CS', ['DERIVED', 'SECONDARY']) # SECONDARY
# Instance Number
ds[0x0020,
0x0013] = DataElement(0x00200013, 'IS',
datasetObj['GeneralImage']['InstanceNumber'])
# Window Center
ds[0x0028, 0x1050] = DataElement(0x00281050, 'DS',
datasetObj['VOILUT']['WindowCenter'])
# Window Width
ds[0x0028, 0x1051] = DataElement(0x00281051, 'DS',
datasetObj['VOILUT']['WindowWidth'])
# # SOP Class UID
# ds[0x0008, 0x0016] = DataElement(0x00080016, 'UI', '1.2.840.10008.5.1.4.1.1.7')
# # SOP Instance UID
# ds[0x0008, 0x0018] = DataElement(0x00080018, 'UI', '1.2.840.1.2.8.236.511020181107154038')
# Bits Allocated
ds[0x0028, 0x0100] = DataElement(0x00280100, 'US', 8)
# Rows
ds[0x0028, 0x0010] = DataElement(0x00280010, 'US', 480)
# Columns
ds[0x0028, 0x0011] = DataElement(0x00280011, 'US', 640)
# Photometric Interpretation
ds[0x0028, 0x0004] = DataElement(0x00280004, 'UI', 'RGB')
# Samples Per Pixel
ds[0x0028, 0x0002] = DataElement(0x00280002, 'US', 0)
# Pixel Aspect Ratio
ds[0x0028, 0x0034] = DataElement(0x00280034, 'IS', ['1', '1'])
# Bits Stored
ds[0x0028, 0x0101] = DataElement(0x00280101, 'US', 8)
# High Bit
ds[0x0028, 0x0102] = DataElement(0x00280102, 'US', 7)
ds.ContentDate = dt.strftime('%Y%m%d')
ds.ContentTime = dt.strftime('%H%M%S.%f') # long format with micro seconds
ds.save_as(outfilename)
# pydicom.write_file(filename, ds)
# print('Load file {} ...'.format(outfilename))
# ds = pydicom.dcmread(outfilename)
# print(str(ds).encode('iso8859').decode('iso8859'))
print("Dicom文件：{} 生成成功！".format(outfilename))
def interpolate_and_wrapup_rs(input_mrcnn_out, input_ct_filelist,
output_rs_filepath_after_ai, model_name):
print('Use RS Template for with model_name = {}'.format(model_name))
if model_name == "MRCNN_Brachy":
#RS_TEMPLATE_FILEPATH = r"./ModelsAndRSTemplates/Brachy/RS_Template/RS.1.2.246.352.71.4.417454940236.267194.20190411111011.dcm"
RS_TEMPLATE_FILEPATH = r"./ModelsAndRSTemplates/NewBrachy/RS_Template/RS.1.2.246.352.71.4.417454940236.267194.20190411111011.dcm"
elif model_name == "MRCNN_Breast":
RS_TEMPLATE_FILEPATH = r"./ModelsAndRSTemplates/Breast/RS_Template/RS.1.2.246.352.71.4.417454940236.244247.20190418132000.dcm"
print("call interpolate_and_wrapup_rs()")
print("with arg input_ct_filelist = {}".format(input_ct_filelist))
print("with arg output_rs_filepath_after_ai = {}".format(
output_rs_filepath_after_ai))
import datetime
import traceback
from pydicom.dataset import Dataset, DataElement
global CT_table
global Thickness_table
CT_table = {}
Thickness_table = {}
rs_fp = pydicom_read_rs_template(input_ct_filelist, RS_TEMPLATE_FILEPATH)
ct_FrameOfReferenceUID = None
ct_StudyInstnaceUID = None
for filepath in input_ct_filelist:
ct_fp = pydicom.read_file(filepath)
ct_FrameOfReferenceUID = ct_fp.FrameOfReferenceUID
ct_StudyInstanceUID = ct_fp.StudyInstanceUID
break
# ================== process_rs() start ===========================
global SOPInstanceUID_list_of_CTs
label_id_mask = input_mrcnn_out
labels = list(label_id_mask.keys())
# =================== Generate Color Mapping ===========================
print("print(labels)=============================================")
print(labels)
# ['bowel', 'Uterus', 'HR-CTV', 'Sigmoid_colon', 'Rectum', 'Bladder', 'Foley']
print("print(labels)=============================================")
# For Brachy template, labels is like ['bowel', 'Uterus', 'HR-CTV', 'Sigmoid_colon', 'Rectum', 'Bladder', 'Foley']
colors = []
# disable code by milochen
#rs_fp = pydicom.read_file(Env.RS_TEMPLATE_FILEPATH)
# Make labels->colors mapping table by information of original template ADD by milochen
# The output result should look like
# ['bowel', 'Uterus', 'HR-CTV', 'Sigmoid_colon', 'Rectum', 'Bladder', 'Foley']
#
# [None, ['255', '255', '128'], ['0', '255', '0'], ['187', '255', '187'], ['128', '64', '64'], ['0', '255', '255'],['0', '150', '0'] ]
for lbl_idx in range(len(labels)):
lblName = labels[lbl_idx]
lblName = lblName.split('RS_')[-1]
# Find the ROI Number that its ROIName equal to label name lblName
isFindMatchedROINumber = False
matchedROINumber = None
for idx in range(len(rs_fp.StructureSetROISequence)):
item = rs_fp.StructureSetROISequence[idx]
if item.ROIName == lblName:
isFindMatchedROINumber = True
matchedROINumber = item.ROINumber
break
if isFindMatchedROINumber == False:
# not Matched , so ignore this stable
print("isFindMatchedROINumber is False when labels[", lbl_idx,
"] = ", lblName)
colors.append(None)
continue
# Find the ROI Display Color that its matchedROINumber equal to ReferencedROINumber
isFindMatchedROIDisplayColor = False
matchedROIDisplayColor = None
#for idx in range(len(rs_fp.ROIContourSequence)):
# item = rs_fp.ROIContourSequence[idx]
for idx, item in enumerate(rs_fp.ROIContourSequence):
if item.ReferencedROINumber == matchedROINumber:
isFindMatchedROIDisplayColor = True
matchedROIDisplayColor = item.ROIDisplayColor
break
if isFindMatchedROIDisplayColor == True:
print('find color work and color = {}'.format(
matchedROIDisplayColor))
else:
print('color is not found')
if isFindMatchedROIDisplayColor == False:
print("isFindMatchedROIDisplayColor is False when labels[",
lbl_idx, "] = ", lblName)
colors.append(None)
continue
colors.append(matchedROIDisplayColor)
print('show colors')
for idx, color in enumerate(colors):
print('idx={} -> {}'.format(idx, color))
# Make color mapping
print('Make color mapping')
colorMapping = {}
for idx in range(len(labels)):
lblName = labels[idx]
lblName = lblName.split('RS_')[-1] # cut off RS_ head if there is
#colorMapping[ labels[idx] ] = colors[idx]
#colorMapping[lblName_ignore_RS_Head] = colors[idx]
colorMapping[lblName] = colors[idx]
#print(labels[idx], "->", colors[idx])
print(lblName, "->", colors[idx])
print("colorMapping Research")
print(labels)
print(colorMapping)
# The output table may look like this
# bowel -> None
# Uterus -> ['255', '255', '128']
# HR - CTV -> ['0', '255', '0']
# Sigmoid_colon -> ['187', '255', '187']
# Rectum -> ['128', '64', '64']
# Bladder -> ['0', '255', '255']
# Foley -> ['0', '150', '0']
# check key for "if key in colorMapping"
# Hightlight problem of code. ROINumber are change, so the order in original rs file is different to output rs file
# The strategy of Sac's original code is to clean all of things and rewrite all.
# =================== Generate Color Mapping ===========================
ct_id = set()
for ct_ids in label_id_mask.values():
ct_id.update(ct_ids.keys())
ct_id = list(ct_id)
ct_id.sort()
# fix bug of SeriesInstnaceUID, it should be a unique SeriesInstnaceUID
# 1.2.246.352.71.2.417454940236.3986270.2019041110054111
import datetime
now = datetime.datetime.now()
theDateForSeriesInstanceUID = now.strftime("%Y%m%d%H%M%S%f")
rs_fp.SeriesInstanceUID = "1.2.246.352.71.2.417454940236.3986270." + theDateForSeriesInstanceUID
rs_fp.ReferencedFrameOfReferenceSequence[
0].FrameOfReferenceUID = ct_FrameOfReferenceUID
rs_fp.ReferencedFrameOfReferenceSequence[0].RTReferencedStudySequence[
0].ReferencedSOPInstanceUID = rs_fp.StudyInstanceUID
rs_fp.ReferencedFrameOfReferenceSequence[0].RTReferencedStudySequence[
0].RTReferencedSeriesSequence[0].ContourImageSequence.clear()
print(
"print_ct_id for check RTReferencedSeriesSequence ContourImageSequence"
)
print(ct_id)
global SOPInstanceUID_list_of_CTs
make_SOPInstnaceUID_list_of_CTs(input_ct_filelist)
# Query here https://imagej.nih.gov/nih-image/download/nih-image_spin-offs/NucMed_Image/DICOM%20Dictionary
#for _id in ct_id:
for _id in SOPInstanceUID_list_of_CTs:
#print("_id = ", _id)
ds = Dataset()
ds[0x0008, 0x1150] = DataElement(
0x00081150, 'UI', 'CT Image Storage') # Referenced SOP Class UID
ds.ReferencedSOPClassUID = '1.2.840.10008.5.1.4.1.1.2'
ds[0x0008, 0x1155] = DataElement(0x00081155, 'UI',
_id) # Referenced SOP Instance UID
rs_fp.ReferencedFrameOfReferenceSequence[0].RTReferencedStudySequence[
0].RTReferencedSeriesSequence[0].ContourImageSequence.append(ds)
rs_fp.StructureSetROISequence.clear()
for i in range(len(labels)):
ds = Dataset()
ds[0x3006, 0x0022] = DataElement(0x30060022, 'IS',
str(i + 1)) # ROI Number
ds[0x3006, 0x0024] = DataElement(
0x30060024, 'UI',
ct_FrameOfReferenceUID) # Referenced Frame of Reference UID
#ds[0x3006, 0x0026] = DataElement(0x30060026, 'LO', labels[i]) # ROI Name
ds[0x3006, 0x0026] = DataElement(0x30060026, 'LO', labels[i].split(
'RS_')[-1]) # ROI Name. if there is RS_ in head, then cut it off
ds[0x3006, 0x0036] = DataElement(0x30060036, 'CS',
'MANUAL') # ROI Generation Algorithm
rs_fp.StructureSetROISequence.append(ds)
rs_fp.ROIContourSequence.clear()
for i in range(len(labels)):
dsss = Dataset()
# changed by milochen
# dsss[0x3006, 0x002a] = DataElement(0x3006002a, 'IS', color)
# Draw color
lblName = labels[i]
if lblName.split('RS_')[-1] in colorMapping.keys():
lblName_ignore_RS_Head = lblName.split('RS_')[-1]
drawColor = colorMapping[lblName_ignore_RS_Head]
print('drawColor = {}'.format(drawColor))
#drawColor = colorMapping[lblName]
#drawColor = colorMapping[lblName]
if drawColor == None:
#dsss[0x3006, 0x002a] = DataElement(0x3006002a, 'IS', [0, 0, 0]) # ROI Display Color
the_color = [255, 255, 0]
if lblName == "RS_PTV":
the_color = [255, 0, 0] # Red
elif lblName == "RS_CTV":
the_color = [0, 255, 0] # Green
dsss[0x3006,
0x002a] = DataElement(0x3006002a, 'IS',
the_color) # ROI Display Color
else:
dsss[0x3006,
0x002a] = DataElement(0x3006002a, 'IS',
drawColor) # ROI Display Color
else:
dsss[0x3006,
0x002a] = DataElement(0x3006002a, 'IS',
[255, 255, 255]) # ROI Display Color
dsss.ContourSequence = []
dsss[0x3006, 0x0084] = DataElement(0x30060084, 'IS',
str(i + 1)) # Referenced ROI Number
rs_fp.ROIContourSequence.append(dsss)
for _id in label_id_mask[labels[i]].keys():
ds = Dataset()
ds[0x0008,
0x1150] = DataElement(0x00081150, 'UI',
'CT Image Storage') # SOP Class UID
ds.ReferencedSOPClassUID = '1.2.840.10008.5.1.4.1.1.2'
ds[0x0008, 0x1155] = DataElement(0x00081155, 'UI',
_id) # SOP Instance UID
NumberOfContourPoints = str(
int(len(label_id_mask[labels[i]][_id]) / 3))
dss = Dataset()
dss.ContourImageSequence = [ds]
dss[0x3006, 0x0042] = DataElement(
0x30060042, 'CS', 'CLOSED_PLANAR') # Contour Geometric Type
# dss[0x3006, 0x0046] = DataElement(0x30060046, 'IS', '664')
dss[0x3006, 0x0046] = DataElement(
0x30060046, 'IS',
NumberOfContourPoints) # Number Of Contour Points
dss[0x3006, 0x0050] = DataElement(
0x30060050, 'DS',
label_id_mask[labels[i]][_id]) # Contour Data
rs_fp.ROIContourSequence[i].ContourSequence.append(dss)
rs_fp.RTROIObservationsSequence.clear()
for i in range(len(labels)):
dss = Dataset()
dss[0x3006, 0x0082] = DataElement(0x30060082, 'IS', str(i + 1))
dss[0x3006, 0x0084] = DataElement(0x30060084, 'IS', str(i + 1))
dss[0x3006, 0x0085] = DataElement(0x30060085, 'SH', labels[i])
dss[0x3006, 0x00a4] = DataElement(0x300600a4, 'CS', '')
dss[0x3006, 0x00a6] = DataElement(0x300600a6, 'PN', '')
rs_fp.RTROIObservationsSequence.append(dss)
#rs_fp.SOPInstanceUID = "{}.{}".format(uid, 1)
rs_fp.SOPInstanceUID = "{}.{}".format(ct_StudyInstanceUID, 1)
rs_fp.ApprovalStatus = "UNAPPROVED"
# ========
#Start to Interpolate
make_some_tables(input_ct_filelist)
for i in range(len(rs_fp.ROIContourSequence)):
try:
item = rs_fp.ROIContourSequence[i].ContourSequence
organ_exist_z = []
# organ_exist_item = []
last_z = 0
for n in range(len(item)):
get_z = float(item[n].ContourData[2])
organ_exist_z.append(get_z)
last_z = get_z
# organ_exist_item.append(n)
# organ_exist_z = np.array(organ_exist_z)
max_Z = max(organ_exist_z)
min_Z = min(organ_exist_z)
print(organ_exist_z)
thickness = Thickness_table[
last_z] # Actually the thickness from any exist z-slice are the same
print('thickness = ', thickness)
organ_table = [
i for i in np.arange(min_Z, max_Z + thickness, thickness)
]
print(organ_table)
for p in range(len(organ_table)):
z = organ_table[p]
# print(z)
# if organ_exist_z.index(z): #z in organ_exist_z
if z not in organ_exist_z: # z in organ_exist_z
# print(z)
to_fill_z = find_nearest(organ_exist_z, z)
to_fill = organ_exist_z.index(to_fill_z)
print(z, '___', to_fill_z, to_fill)
# cpyItem = copy.deepcopy(rs_fp.ROIContourSequence[i].ContourSequence[to_fill])
to_fill_data = rs_fp.ROIContourSequence[i].ContourSequence[
to_fill]
deepcopy_data = copy.deepcopy(to_fill_data)
# rs_fp.ROIContourSequence[i].ContourSequence.append(to_fill_data)
rs_fp.ROIContourSequence[i].ContourSequence.append(
deepcopy_data)
# item.ContourSequence.....append(fill_z)
# edit
#rs_fp.ROIContourSequence[i].ContourSequence[-1].ContourImageSequence[0].ReferencedSOPInstanceUID = CT_table[to_fill_z]
rs_fp.ROIContourSequence[i].ContourSequence[
-1].ContourImageSequence[
0].ReferencedSOPInstanceUID = CT_table[z]
rs_fp.ROIContourSequence[i].ContourSequence[
-1].ContourData[2::3] = [z] * (int(
len(rs_fp.ROIContourSequence[i].
ContourSequence[-1].ContourData) / 3))
except Exception as e:
print("Exception")
traceback.print_tb(e.__traceback__)
#thickness = Thickness_table[last_z]
print("last_z = ", last_z)
print("Thickness_table = ", Thickness_table)
#print(e)
#set StructureSetLabel from RALCT_20190411 to RALCT_20190641_AI. The 20190614 is
studyDate = lastGetDateFromCT
# Change to 16 bytes because Varian cannot allow 17 bytes
# rs_fp.StructureSetLabel = "RALCT_{}_AI".format(studyDate)
rs_fp.StructureSetLabel = "RALCT_{}_A".format(studyDate)
print("write_file for ", output_rs_filepath_after_ai)
pydicom.write_file(output_rs_filepath_after_ai, rs_fp)
return None
def test_anonymise_dataset_and_all_is_anonymised_functions():
# Create dict with one instance of every identifying keyword and
# run basic anonymisation tests
non_anon_dict = dict.fromkeys(IDENTIFYING_KEYWORDS)
for key in non_anon_dict:
non_anon_dict[key] = _get_non_anonymous_replacement_value(key)
ds = dicom_dataset_from_dict(non_anon_dict)
_check_is_anonymised_dataset_file_and_dir(ds, anon_is_expected=False)
ds_anon = anonymise_dataset(ds)
_check_is_anonymised_dataset_file_and_dir(ds_anon, anon_is_expected=True)
# Test anonymisation (and check thereof) for each identifying
# element individually.
for elem in ds_anon.iterall():
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,
anon_is_expected=False)
ds_single_anon = anonymise_dataset(ds_single_non_anon_value)
_check_is_anonymised_dataset_file_and_dir(ds_single_anon,
anon_is_expected=True)
# Test correct handling of private tags
ds_anon.add(DataElement(0x0043102b, 'SS', [4, 4, 0, 0]))
_check_is_anonymised_dataset_file_and_dir(ds_anon,
anon_is_expected=False,
ignore_private_tags=False)
_check_is_anonymised_dataset_file_and_dir(ds_anon,
anon_is_expected=True,
ignore_private_tags=True)
ds_anon.remove_private_tags()
_check_is_anonymised_dataset_file_and_dir(ds_anon,
anon_is_expected=True,
ignore_private_tags=False)
# Test blank anonymisation
# # Sanity check
_check_is_anonymised_dataset_file_and_dir(ds, anon_is_expected=False)
ds_anon_blank = anonymise_dataset(ds, replace_values=False)
_check_is_anonymised_dataset_file_and_dir(ds_anon_blank,
anon_is_expected=True)
# Test handling of unknown tags by removing PatientName from
# baseline dict
patient_name_tag = tag_for_keyword('PatientName')
try:
patient_name = BaselineDicomDictionary.pop(patient_name_tag)
with pytest.raises(ValueError) as e_info:
anonymise_dataset(ds)
assert str(e_info).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,
anon_is_expected=True)
with pytest.raises(AttributeError) as e_info:
ds_anon_delete_unknown.PatientName
assert str(e_info).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,
anon_is_expected=True)
assert patient_name_tag in ds_anon_ignore_unknown
finally:
BaselineDicomDictionary.setdefault(patient_name_tag, patient_name)
# Test copy_dataset=False:
anonymise_dataset(ds, copy_dataset=False)
assert is_anonymised_dataset(ds)
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 = Dataset()
for keyword in IDENTIFYING_KEYWORDS:
# Ignore file meta elements for now
tag = hex(tag_for_keyword(keyword))
if 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(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 = tag_for_keyword("PatientName")
try:
patient_name = 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
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:
BASELINE_DICOM_DICT.setdefault(patient_name_tag, patient_name)
# Test copy_dataset=False:
anonymise_dataset(ds, copy_dataset=False)
assert is_anonymised_dataset(ds)