def test(path_data='', parameters=''):
verbose = 0
dice_threshold = 0.9
add_path_for_template = False # if absolute path or no path to template is provided, then path to data should not be added.
# initializations
dice_template2anat = float('NaN')
dice_anat2template = float('NaN')
output = ''
if not parameters:
parameters = '-i t2/t2.nii.gz -l t2/labels.nii.gz -s t2/t2_seg.nii.gz ' \
'-param step=1,type=seg,algo=centermassrot,metric=MeanSquares:step=2,type=seg,algo=bsplinesyn,iter=5,metric=MeanSquares ' \
'-t template/ -r 0'
add_path_for_template = True # in this case, path to data should be added
parser = sct_register_to_template.get_parser()
dict_param = parser.parse(parameters.split(), check_file_exist=False)
if add_path_for_template:
dict_param_with_path = parser.add_path_to_file(deepcopy(dict_param),
path_data,
input_file=True)
else:
dict_param_with_path = parser.add_path_to_file(deepcopy(dict_param),
path_data,
input_file=True,
do_not_add_path=['-t'])
param_with_path = parser.dictionary_to_string(dict_param_with_path)
# Check if input files exist
if not (os.path.isfile(dict_param_with_path['-i'])
and os.path.isfile(dict_param_with_path['-l'])
and os.path.isfile(dict_param_with_path['-s'])):
status = 200
output = 'ERROR: the file(s) provided to test function do not exist in folder: ' + path_data
return status, output, DataFrame(data={
'status': int(status),
'output': output
},
index=[path_data])
# return status, output, DataFrame(
# data={'status': status, 'output': output,
# 'dice_template2anat': float('nan'), 'dice_anat2template': float('nan')},
# index=[path_data])
# if template is not specified, use default
# if not os.path.isdir(dict_param_with_path['-t']):
# status, path_sct = commands.getstatusoutput('echo $SCT_DIR')
# dict_param_with_path['-t'] = path_sct + default_template
# param_with_path = parser.dictionary_to_string(dict_param_with_path)
# get contrast folder from -i option.
# We suppose we can extract it as the first object when spliting with '/' delimiter.
contrast_folder = ''
input_filename = ''
if dict_param['-i'][0] == '/':
dict_param['-i'] = dict_param['-i'][1:]
input_split = dict_param['-i'].split('/')
if len(input_split) == 2:
contrast_folder = input_split[0] + '/'
input_filename = input_split[1]
else:
input_filename = input_split[0]
if not contrast_folder: # if no contrast folder, send error.
status = 201
output = 'ERROR: when extracting the contrast folder from input file in command line: ' + dict_param[
'-i'] + ' for ' + path_data
return status, output, DataFrame(data={
'status': int(status),
'output': output
},
index=[path_data])
# return status, output, DataFrame(
# data={'status': status, 'output': output, 'dice_template2anat': float('nan'), 'dice_anat2template': float('nan')}, index=[path_data])
# create output path
# TODO: create function for that
import time, random
subject_folder = path_data.split('/')
if subject_folder[-1] == '' and len(subject_folder) > 1:
subject_folder = subject_folder[-2]
else:
subject_folder = subject_folder[-1]
path_output = sct.slash_at_the_end(
'sct_register_to_template_' + subject_folder + '_' +
time.strftime("%y%m%d%H%M%S") + '_' + str(random.randint(1, 1000000)),
slash=1)
param_with_path += ' -ofolder ' + path_output
sct.create_folder(path_output)
# log file
# TODO: create function for that
import sys
fname_log = path_output + 'output.log'
sct.pause_stream_logger()
file_handler = sct.add_file_handler_to_logger(filename=fname_log,
mode='w',
log_format="%(message)s")
#
# stdout_log = file(fname_log, 'w')
# redirect to log file
# stdout_orig = sys.stdout
# sys.stdout = stdout_log
cmd = 'sct_register_to_template ' + param_with_path
output += '\n====================================================================================================\n' + cmd + '\n====================================================================================================\n\n' # copy command
time_start = time.time()
try:
status, o = sct.run(cmd, verbose)
except:
status, o = 1, 'ERROR: Function crashed!'
output += o
duration = time.time() - time_start
# if command ran without error, test integrity
if status == 0:
# get filename_template_seg
fname_template_seg = get_file_label(
sct.slash_at_the_end(dict_param_with_path['-t'], 1) + 'template/',
'spinal cord',
output='filewithpath')
# apply transformation to binary mask: template --> anat
sct.run(
'sct_apply_transfo -i ' + fname_template_seg + ' -d ' +
dict_param_with_path['-s'] + ' -w ' + path_output +
'warp_template2anat.nii.gz' + ' -o ' + path_output +
'test_template2anat.nii.gz -x nn', verbose)
# apply transformation to binary mask: anat --> template
sct.run(
'sct_apply_transfo -i ' + dict_param_with_path['-s'] + ' -d ' +
fname_template_seg + ' -w ' + path_output +
'warp_anat2template.nii.gz' + ' -o ' + path_output +
'test_anat2template.nii.gz -x nn', verbose)
# compute dice coefficient between template segmentation warped into anat and segmentation from anat
cmd = 'sct_dice_coefficient -i ' + dict_param_with_path[
'-s'] + ' -d ' + path_output + 'test_template2anat.nii.gz'
status1, output1 = sct.run(cmd, verbose)
# parse output and compare to acceptable threshold
dice_template2anat = float(
output1.split('3D Dice coefficient = ')[1].split('\n')[0])
if dice_template2anat < dice_threshold:
status1 = 99
# compute dice coefficient between segmentation from anat warped into template and template segmentation
# N.B. here we use -bmax because the FOV of the anat is smaller than the template
cmd = 'sct_dice_coefficient -i ' + fname_template_seg + ' -d ' + path_output + 'test_anat2template.nii.gz -bmax 1'
status2, output2 = sct.run(cmd, verbose)
# parse output and compare to acceptable threshold
dice_anat2template = float(
output2.split('3D Dice coefficient = ')[1].split('\n')[0])
if dice_anat2template < dice_threshold:
status2 = 99
# check if at least one integrity status was equal to 99
if status1 == 99 or status2 == 99:
status = 99
# concatenate outputs
output = output + output1 + output2
# transform results into Pandas structure
results = DataFrame(data={
'status': int(status),
'output': output,
'dice_template2anat': dice_template2anat,
'dice_anat2template': dice_anat2template,
'duration [s]': duration
},
index=[path_data])
sct.log.info(output)
sct.remove_handler(file_handler)
sct.start_stream_logger()
return status, output, results
def test(path_data='', parameters=''):
verbose = 0
dice_threshold = 0.9
add_path_for_template = False # if absolute path or no path to template is provided, then path to data should not be added.
# initializations
dice_template2anat = float('NaN')
dice_anat2template = float('NaN')
output = ''
if not parameters:
parameters = '-i t2/t2.nii.gz -l t2/labels.nii.gz -s t2/t2_seg.nii.gz ' \
'-param step=1,type=seg,algo=centermassrot,metric=MeanSquares:step=2,type=seg,algo=bsplinesyn,iter=5,metric=MeanSquares ' \
'-t template/ -r 0'
add_path_for_template = True # in this case, path to data should be added
parser = sct_register_to_template.get_parser()
dict_param = parser.parse(parameters.split(), check_file_exist=False)
if add_path_for_template:
dict_param_with_path = parser.add_path_to_file(deepcopy(dict_param), path_data, input_file=True)
else:
dict_param_with_path = parser.add_path_to_file(deepcopy(dict_param), path_data, input_file=True, do_not_add_path=['-t'])
param_with_path = parser.dictionary_to_string(dict_param_with_path)
# Check if input files exist
if not (os.path.isfile(dict_param_with_path['-i']) and
os.path.isfile(dict_param_with_path['-l']) and
os.path.isfile(dict_param_with_path['-s'])):
status = 200
output = 'ERROR: the file(s) provided to test function do not exist in folder: ' + path_data
return status, output, DataFrame(data={'status': int(status), 'output': output}, index=[path_data])
# return status, output, DataFrame(
# data={'status': status, 'output': output,
# 'dice_template2anat': float('nan'), 'dice_anat2template': float('nan')},
# index=[path_data])
# if template is not specified, use default
# if not os.path.isdir(dict_param_with_path['-t']):
# status, path_sct = commands.getstatusoutput('echo $SCT_DIR')
# dict_param_with_path['-t'] = path_sct + default_template
# param_with_path = parser.dictionary_to_string(dict_param_with_path)
# get contrast folder from -i option.
# We suppose we can extract it as the first object when spliting with '/' delimiter.
contrast_folder = ''
input_filename = ''
if dict_param['-i'][0] == '/':
dict_param['-i'] = dict_param['-i'][1:]
input_split = dict_param['-i'].split('/')
if len(input_split) == 2:
contrast_folder = input_split[0] + '/'
input_filename = input_split[1]
else:
input_filename = input_split[0]
if not contrast_folder: # if no contrast folder, send error.
status = 201
output = 'ERROR: when extracting the contrast folder from input file in command line: ' + dict_param['-i'] + ' for ' + path_data
return status, output, DataFrame(data={'status': int(status), 'output': output}, index=[path_data])
# return status, output, DataFrame(
# data={'status': status, 'output': output, 'dice_template2anat': float('nan'), 'dice_anat2template': float('nan')}, index=[path_data])
import time, random
subject_folder = path_data.split('/')
if subject_folder[-1] == '' and len(subject_folder) > 1:
subject_folder = subject_folder[-2]
else:
subject_folder = subject_folder[-1]
path_output = sct.slash_at_the_end('sct_register_to_template_' + subject_folder + '_' + time.strftime("%y%m%d%H%M%S") + '_'+str(random.randint(1, 1000000)), slash=1)
param_with_path += ' -ofolder ' + path_output
cmd = 'sct_register_to_template ' + param_with_path
output += '\n====================================================================================================\n'+cmd+'\n====================================================================================================\n\n' # copy command
time_start = time.time()
try:
status, o = sct.run(cmd, verbose)
except:
status, o = 1, 'ERROR: Function crashed!'
output += o
duration = time.time() - time_start
# if command ran without error, test integrity
if status == 0:
# get filename_template_seg
fname_template_seg = get_file_label(sct.slash_at_the_end(dict_param_with_path['-t'], 1) + 'template/', 'spinal cord', output='filewithpath')
# apply transformation to binary mask: template --> anat
sct.run(
'sct_apply_transfo -i ' + fname_template_seg +
' -d ' + dict_param_with_path['-s'] +
' -w ' + path_output + 'warp_template2anat.nii.gz' +
' -o ' + path_output + 'test_template2anat.nii.gz -x nn', verbose)
# apply transformation to binary mask: anat --> template
sct.run(
'sct_apply_transfo -i ' + dict_param_with_path['-s'] +
' -d ' + fname_template_seg +
' -w ' + path_output + 'warp_anat2template.nii.gz' +
' -o ' + path_output + 'test_anat2template.nii.gz -x nn', verbose)
# compute dice coefficient between template segmentation warped into anat and segmentation from anat
cmd = 'sct_dice_coefficient -i ' + dict_param_with_path['-s'] + ' -d ' + path_output + 'test_template2anat.nii.gz'
status1, output1 = sct.run(cmd, verbose)
# parse output and compare to acceptable threshold
dice_template2anat = float(output1.split('3D Dice coefficient = ')[1].split('\n')[0])
if dice_template2anat < dice_threshold:
status1 = 99
# compute dice coefficient between segmentation from anat warped into template and template segmentation
# N.B. here we use -bmax because the FOV of the anat is smaller than the template
cmd = 'sct_dice_coefficient -i ' + fname_template_seg + ' -d ' + path_output + 'test_anat2template.nii.gz -bmax 1'
status2, output2 = sct.run(cmd, verbose)
# parse output and compare to acceptable threshold
dice_anat2template = float(output2.split('3D Dice coefficient = ')[1].split('\n')[0])
if dice_anat2template < dice_threshold:
status2 = 99
# check if at least one integrity status was equal to 99
if status1 == 99 or status2 == 99:
status = 99
# concatenate outputs
output = output + output1 + output2
# transform results into Pandas structure
results = DataFrame(data={'status': int(status), 'output': output, 'dice_template2anat': dice_template2anat, 'dice_anat2template': dice_anat2template, 'duration [s]': duration}, index=[path_data])
return status, output, results
def test(path_data='', parameters=''):
verbose = 0
filename_template = 'template/MNI-Poly-AMU_cord.nii.gz'
dice_threshold = 0.85
if not parameters:
parameters = '-i t2/t2.nii.gz -l t2/labels.nii.gz -s t2/t2_seg.nii.gz ' \
'-param step=1,type=seg,algo=slicereg,metric=MeanSquares,iter=5:step=2,type=seg,algo=bsplinesyn,iter=3,metric=MI:step=3,iter=0 ' \
'-t template/ -r 0'
parser = sct_register_to_template.get_parser()
dict_param = parser.parse(parameters.split(), check_file_exist=False)
dict_param_with_path = parser.add_path_to_file(dict_param, path_data, input_file=True)
param_with_path = parser.dictionary_to_string(dict_param_with_path)
# Check if input files exist
if not (os.path.isfile(dict_param_with_path['-i']) and
os.path.isfile(dict_param_with_path['-l']) and
os.path.isfile(dict_param_with_path['-s'])):
status = 200
output = 'ERROR: the file(s) provided to test function do not exist in folder: ' + path_data
return status, output, DataFrame(
data={'status': status, 'output': output,
'dice_template2anat': float('nan'), 'dice_anat2template': float('nan')},
index=[path_data])
if not os.path.isdir(dict_param_with_path['-t']):
status, path_sct = commands.getstatusoutput('echo $SCT_DIR')
dict_param_with_path['-t'] = path_sct + '/data/template/'
param_with_path = parser.dictionary_to_string(dict_param_with_path)
# get contrast folder from -i option.
# We suppose we can extract it as the first object when spliting with '/' delimiter.
contrast_folder = ''
input_split = dict_param['-i'].split('/')
if input_split[0]:
contrast_folder = input_split[0] + '/'
else:
contrast_folder = input_split[1] + '/'
if not contrast_folder: # if no contrast folder, send error.
status = 201
output = 'ERROR: when extracting the contrast folder from input file in command line: ' + dict_param['-i'] + ' for ' + path_data
return status, output, DataFrame(
data={'status': status, 'output': output, 'mse': float('nan'), 'dist_max': float('nan')}, index=[path_data])
import time, random
subject_folder = path_data.split('/')
if subject_folder[-1] == '' and len(subject_folder) > 1:
subject_folder = subject_folder[-2]
else:
subject_folder = subject_folder[-1]
path_output = sct.slash_at_the_end('sct_register_to_template_' + subject_folder + '_' + time.strftime("%y%m%d%H%M%S") + '_'+str(random.randint(1, 1000000)), slash=1)
param_with_path += ' -ofolder ' + path_output
cmd = 'sct_register_to_template ' + param_with_path
output = '\n====================================================================================================\n'+cmd+'\n====================================================================================================\n\n' # copy command
time_start = time.time()
status, o = sct.run(cmd, verbose)
output += o
duration = time.time() - time_start
# if command ran without error, test integrity
if status == 0:
# apply transformation to binary mask: template --> anat
sct.run(
'sct_apply_transfo -i ' + dict_param_with_path['-t'] + filename_template +
' -d ' + dict_param_with_path['-s'] +
' -w ' + path_output + 'warp_template2anat.nii.gz' +
' -o ' + path_output + 'test_template2anat.nii.gz -x nn', verbose)
# apply transformation to binary mask: anat --> template
sct.run(
'sct_apply_transfo -i ' + dict_param_with_path['-s'] +
' -d ' + dict_param_with_path['-t'] + filename_template +
' -w ' + path_output + 'warp_anat2template.nii.gz' +
' -o ' + path_output + 'test_anat2template.nii.gz -x nn', verbose)
# compute dice coefficient between template segmentation warped into anat and segmentation from anat
cmd = 'sct_dice_coefficient -i ' + dict_param_with_path['-s'] + ' -d ' + path_output + 'test_template2anat.nii.gz'
status1, output1 = sct.run(cmd, verbose)
# parse output and compare to acceptable threshold
dice_template2anat = float(output1.split('3D Dice coefficient = ')[1].split('\n')[0])
if dice_template2anat < dice_threshold:
status1 = 99
# compute dice coefficient between segmentation from anat warped into template and template segmentation
# N.B. here we use -bmax because the FOV of the anat is smaller than the template
cmd = 'sct_dice_coefficient -i ' + dict_param_with_path['-t'] + filename_template + ' -d ' + path_output + 'test_anat2template.nii.gz -bmax 1'
status2, output2 = sct.run(cmd, verbose)
# parse output and compare to acceptable threshold
dice_anat2template = float(output2.split('3D Dice coefficient = ')[1].split('\n')[0])
if dice_anat2template < dice_threshold:
status2 = 99
# check if at least one integrity status was equal to 99
if status1 == 99 or status2 == 99:
status = 99
# concatenate outputs
output = output + output1 + output2
# transform results into Pandas structure
results = DataFrame(data={'status': status, 'output': output, 'dice_template2anat': dice_template2anat, 'dice_anat2template': dice_anat2template, 'duration [s]': duration}, index=[path_data])
return status, output, results
