def main():
"""Main function."""
parser = get_parser()
args = parser.parse_args(args=None if sys.argv[1:] else ['--help'])
fname_image = os.path.abspath(args.i)
contrast_type = args.c
ctr_algo = args.centerline
if args.brain is None:
if contrast_type in ['t2s', 'dwi']:
brain_bool = False
if contrast_type in ['t1', 't2']:
brain_bool = True
else:
brain_bool = bool(args.brain)
kernel_size = args.kernel
if kernel_size == '3d' and contrast_type == 'dwi':
kernel_size = '2d'
sct.printv('3D kernel model for dwi contrast is not available. 2D kernel model is used instead.',
type="warning")
if ctr_algo == 'file' and args.file_centerline is None:
sct.printv('Please use the flag -file_centerline to indicate the centerline filename.', 1, 'warning')
sys.exit(1)
if args.file_centerline is not None:
manual_centerline_fname = args.file_centerline
ctr_algo = 'file'
else:
manual_centerline_fname = None
remove_temp_files = args.r
verbose = args.v
sct.init_sct(log_level=verbose, update=True) # Update log level
path_qc = args.qc
qc_dataset = args.qc_dataset
qc_subject = args.qc_subject
output_folder = args.ofolder
algo_config_stg = '\nMethod:'
algo_config_stg += '\n\tCenterline algorithm: ' + str(ctr_algo)
algo_config_stg += '\n\tAssumes brain section included in the image: ' + str(brain_bool)
algo_config_stg += '\n\tDimension of the segmentation kernel convolutions: ' + kernel_size + '\n'
sct.printv(algo_config_stg)
# Segment image
from spinalcordtoolbox.image import Image
from spinalcordtoolbox.deepseg_sc.core import deep_segmentation_spinalcord
from spinalcordtoolbox.reports.qc import generate_qc
im_image = Image(fname_image)
# note: below we pass im_image.copy() otherwise the field absolutepath becomes None after execution of this function
im_seg, im_image_RPI_upsamp, im_seg_RPI_upsamp, im_labels_viewer, im_ctr = \
deep_segmentation_spinalcord(im_image.copy(), contrast_type, ctr_algo=ctr_algo,
ctr_file=manual_centerline_fname, brain_bool=brain_bool, kernel_size=kernel_size,
remove_temp_files=remove_temp_files, verbose=verbose)
# Save segmentation
fname_seg = os.path.abspath(os.path.join(output_folder, sct.extract_fname(fname_image)[1] + '_seg' +
sct.extract_fname(fname_image)[2]))
# copy q/sform from input image to output segmentation
im_seg.copy_qform_from_ref(im_image)
im_seg.save(fname_seg)
if ctr_algo == 'viewer':
# Save labels
fname_labels = os.path.abspath(os.path.join(output_folder, sct.extract_fname(fname_image)[1] + '_labels-centerline' +
sct.extract_fname(fname_image)[2]))
im_labels_viewer.save(fname_labels)
if verbose == 2:
# Save ctr
fname_ctr = os.path.abspath(os.path.join(output_folder, sct.extract_fname(fname_image)[1] + '_centerline' +
sct.extract_fname(fname_image)[2]))
im_ctr.save(fname_ctr)
if path_qc is not None:
generate_qc(fname_image, fname_seg=fname_seg, args=sys.argv[1:], path_qc=os.path.abspath(path_qc),
dataset=qc_dataset, subject=qc_subject, process='sct_deepseg_sc')
sct.display_viewer_syntax([fname_image, fname_seg], colormaps=['gray', 'red'], opacities=['', '0.7'])
def interpolate_im_to_ref(im_input, im_input_sc, new_res=0.3, sq_size_size_mm=22.5, interpolation_mode=3):
nx, ny, nz, nt, px, py, pz, pt = im_input.dim
im_input_sc = im_input_sc.copy()
im_input = im_input.copy()
# keep only spacing and origin in qform to avoid rotation issues
input_qform = im_input.hdr.get_qform()
for i in range(4):
for j in range(4):
if i != j and j != 3:
input_qform[i, j] = 0
im_input.hdr.set_qform(input_qform)
im_input.hdr.set_sform(input_qform)
im_input_sc.hdr = im_input.hdr
sq_size = int(sq_size_size_mm / new_res)
# create a reference image : square of ones
im_ref = Image(np.ones((sq_size, sq_size, 1), dtype=np.int), dim=(sq_size, sq_size, 1, 0, new_res, new_res, pz, 0), orientation='RPI')
# copy input qform matrix to reference image
im_ref.hdr.set_qform(im_input.hdr.get_qform())
im_ref.hdr.set_sform(im_input.hdr.get_sform())
# set correct header to reference image
im_ref.hdr.set_data_shape((sq_size, sq_size, 1))
im_ref.hdr.set_zooms((new_res, new_res, pz))
# save image to set orientation to RPI (not properly done at the creation of the image)
fname_ref = 'im_ref.nii.gz'
im_ref.save(fname_ref).change_orientation("RPI")
# set header origin to zero to get physical coordinates of the center of the square
im_ref.hdr.as_analyze_map()['qoffset_x'] = 0
im_ref.hdr.as_analyze_map()['qoffset_y'] = 0
im_ref.hdr.as_analyze_map()['qoffset_z'] = 0
im_ref.hdr.set_sform(im_ref.hdr.get_qform())
im_ref.hdr.set_qform(im_ref.hdr.get_qform())
[[x_square_center_phys, y_square_center_phys, z_square_center_phys]] = im_ref.transfo_pix2phys(coordi=[[int(sq_size / 2), int(sq_size / 2), 0]])
list_interpolate_images = []
# iterate on z dimension of input image
for iz in range(nz):
# copy reference image: one reference image per slice
im_ref_slice_iz = im_ref.copy()
# get center of mass of SC for slice iz
x_seg, y_seg = (im_input_sc.data[:, :, iz] > 0).nonzero()
x_center, y_center = np.mean(x_seg), np.mean(y_seg)
[[x_center_phys, y_center_phys, z_center_phys]] = im_input_sc.transfo_pix2phys(coordi=[[x_center, y_center, iz]])
# center reference image on SC for slice iz
im_ref_slice_iz.hdr.as_analyze_map()['qoffset_x'] = x_center_phys - x_square_center_phys
im_ref_slice_iz.hdr.as_analyze_map()['qoffset_y'] = y_center_phys - y_square_center_phys
im_ref_slice_iz.hdr.as_analyze_map()['qoffset_z'] = z_center_phys
im_ref_slice_iz.hdr.set_sform(im_ref_slice_iz.hdr.get_qform())
im_ref_slice_iz.hdr.set_qform(im_ref_slice_iz.hdr.get_qform())
# interpolate input image to reference image
im_input_interpolate_iz = im_input.interpolate_from_image(im_ref_slice_iz, interpolation_mode=interpolation_mode, border='nearest')
# reshape data to 2D if needed
if len(im_input_interpolate_iz.data.shape) == 3:
im_input_interpolate_iz.data = im_input_interpolate_iz.data.reshape(im_input_interpolate_iz.data.shape[:-1])
# add slice to list
list_interpolate_images.append(im_input_interpolate_iz)
return list_interpolate_images
def main(args=None):
# Initialization
param = Param()
start_time = time.time()
parser = get_parser()
arguments = parser.parse_args(args=None if sys.argv[1:] else ['--help'])
fname_anat = arguments.i
fname_centerline = arguments.s
param.algo_fitting = arguments.algo_fitting
if arguments.smooth is not None:
sigma = arguments.smooth
remove_temp_files = arguments.r
verbose = int(arguments.v)
init_sct(log_level=verbose, update=True) # Update log level
# Display arguments
printv('\nCheck input arguments...')
printv(' Volume to smooth .................. ' + fname_anat)
printv(' Centerline ........................ ' + fname_centerline)
printv(' Sigma (mm) ........................ ' + str(sigma))
printv(' Verbose ........................... ' + str(verbose))
# Check that input is 3D:
nx, ny, nz, nt, px, py, pz, pt = Image(fname_anat).dim
dim = 4 # by default, will be adjusted later
if nt == 1:
dim = 3
if nz == 1:
dim = 2
if dim == 4:
printv('WARNING: the input image is 4D, please split your image to 3D before smoothing spinalcord using :\n'
'sct_image -i ' + fname_anat + ' -split t -o ' + fname_anat, verbose, 'warning')
printv('4D images not supported, aborting ...', verbose, 'error')
# Extract path/file/extension
path_anat, file_anat, ext_anat = extract_fname(fname_anat)
path_centerline, file_centerline, ext_centerline = extract_fname(fname_centerline)
path_tmp = tmp_create(basename="smooth_spinalcord")
# Copying input data to tmp folder
printv('\nCopying input data to tmp folder and convert to nii...', verbose)
copy(fname_anat, os.path.join(path_tmp, "anat" + ext_anat))
copy(fname_centerline, os.path.join(path_tmp, "centerline" + ext_centerline))
# go to tmp folder
curdir = os.getcwd()
os.chdir(path_tmp)
# convert to nii format
convert('anat' + ext_anat, 'anat.nii')
convert('centerline' + ext_centerline, 'centerline.nii')
# Change orientation of the input image into RPI
printv('\nOrient input volume to RPI orientation...')
fname_anat_rpi = Image("anat.nii") \
.change_orientation("RPI", generate_path=True) \
.save() \
.absolutepath
# Change orientation of the input image into RPI
printv('\nOrient centerline to RPI orientation...')
fname_centerline_rpi = Image("centerline.nii") \
.change_orientation("RPI", generate_path=True) \
.save() \
.absolutepath
# Straighten the spinal cord
# straighten segmentation
printv('\nStraighten the spinal cord using centerline/segmentation...', verbose)
cache_sig = cache_signature(input_files=[fname_anat_rpi, fname_centerline_rpi],
input_params={"x": "spline"})
cachefile = os.path.join(curdir, "straightening.cache")
if cache_valid(cachefile, cache_sig) and os.path.isfile(os.path.join(curdir, 'warp_curve2straight.nii.gz')) and os.path.isfile(os.path.join(curdir, 'warp_straight2curve.nii.gz')) and os.path.isfile(os.path.join(curdir, 'straight_ref.nii.gz')):
# if they exist, copy them into current folder
printv('Reusing existing warping field which seems to be valid', verbose, 'warning')
copy(os.path.join(curdir, 'warp_curve2straight.nii.gz'), 'warp_curve2straight.nii.gz')
copy(os.path.join(curdir, 'warp_straight2curve.nii.gz'), 'warp_straight2curve.nii.gz')
copy(os.path.join(curdir, 'straight_ref.nii.gz'), 'straight_ref.nii.gz')
# apply straightening
run_proc(['sct_apply_transfo', '-i', fname_anat_rpi, '-w', 'warp_curve2straight.nii.gz', '-d', 'straight_ref.nii.gz', '-o', 'anat_rpi_straight.nii', '-x', 'spline'], verbose)
else:
run_proc(['sct_straighten_spinalcord', '-i', fname_anat_rpi, '-o', 'anat_rpi_straight.nii', '-s', fname_centerline_rpi, '-x', 'spline', '-param', 'algo_fitting=' + param.algo_fitting], verbose)
cache_save(cachefile, cache_sig)
# move warping fields locally (to use caching next time)
copy('warp_curve2straight.nii.gz', os.path.join(curdir, 'warp_curve2straight.nii.gz'))
copy('warp_straight2curve.nii.gz', os.path.join(curdir, 'warp_straight2curve.nii.gz'))
# Smooth the straightened image along z
printv('\nSmooth the straightened image...')
sigma_smooth = ",".join([str(i) for i in sigma])
sct_maths.main(args=['-i', 'anat_rpi_straight.nii',
'-smooth', sigma_smooth,
'-o', 'anat_rpi_straight_smooth.nii',
'-v', '0'])
# Apply the reversed warping field to get back the curved spinal cord
printv('\nApply the reversed warping field to get back the curved spinal cord...')
run_proc(['sct_apply_transfo', '-i', 'anat_rpi_straight_smooth.nii', '-o', 'anat_rpi_straight_smooth_curved.nii', '-d', 'anat.nii', '-w', 'warp_straight2curve.nii.gz', '-x', 'spline'], verbose)
# replace zeroed voxels by original image (issue #937)
printv('\nReplace zeroed voxels by original image...', verbose)
nii_smooth = Image('anat_rpi_straight_smooth_curved.nii')
data_smooth = nii_smooth.data
data_input = Image('anat.nii').data
indzero = np.where(data_smooth == 0)
data_smooth[indzero] = data_input[indzero]
nii_smooth.data = data_smooth
nii_smooth.save('anat_rpi_straight_smooth_curved_nonzero.nii')
# come back
os.chdir(curdir)
# Generate output file
printv('\nGenerate output file...')
generate_output_file(os.path.join(path_tmp, "anat_rpi_straight_smooth_curved_nonzero.nii"),
file_anat + '_smooth' + ext_anat)
# Remove temporary files
if remove_temp_files == 1:
printv('\nRemove temporary files...')
rmtree(path_tmp)
# Display elapsed time
elapsed_time = time.time() - start_time
printv('\nFinished! Elapsed time: ' + str(int(np.round(elapsed_time))) + 's\n')
display_viewer_syntax([file_anat, file_anat + '_smooth'], verbose=verbose)
def _orient(self, fname, orientation):
return Image(fname).change_orientation(orientation).save(fname,
mutable=True)
def main(args=None):
"""
Main function
:param args:
:return:
"""
# initializations
output_type = None
dim_list = ['x', 'y', 'z', 't']
# Get parser args
if args is None:
args = None if sys.argv[1:] else ['--help']
parser = get_parser()
arguments = parser.parse_args(args=args)
fname_in = arguments.i
n_in = len(fname_in)
verbose = arguments.v
sct.init_sct(log_level=verbose, update=True) # Update log level
if arguments.o is not None:
fname_out = arguments.o
else:
fname_out = None
# Run command
# Arguments are sorted alphabetically (not according to the usage order)
if arguments.concat is not None:
dim = arguments.concat
assert dim in dim_list
dim = dim_list.index(dim)
im_out = [concat_data(fname_in, dim)] # TODO: adapt to fname_in
elif arguments.copy_header is not None:
im_in = Image(fname_in[0])
im_dest = Image(arguments.copy_header)
im_dest_new = im_in.copy()
im_dest_new.data = im_dest.data.copy()
# im_dest.header = im_in.header
im_dest_new.absolutepath = im_dest.absolutepath
im_out = [im_dest_new]
fname_out = arguments.copy_header
elif arguments.display_warp:
im_in = fname_in[0]
visualize_warp(im_in, fname_grid=None, step=3, rm_tmp=True)
im_out = None
elif arguments.getorient:
im_in = Image(fname_in[0])
orient = im_in.orientation
im_out = None
elif arguments.keep_vol is not None:
index_vol = (arguments.keep_vol).split(',')
for iindex_vol, vol in enumerate(index_vol):
index_vol[iindex_vol] = int(vol)
im_in = Image(fname_in[0])
im_out = [remove_vol(im_in, index_vol, todo='keep')]
elif arguments.mcs:
im_in = Image(fname_in[0])
if n_in != 1:
sct.printv(parser.error('ERROR: -mcs need only one input'))
if len(im_in.data.shape) != 5:
sct.printv(
parser.error(
'ERROR: -mcs input need to be a multi-component image'))
im_out = multicomponent_split(im_in)
elif arguments.omc:
im_ref = Image(fname_in[0])
for fname in fname_in:
im = Image(fname)
if im.data.shape != im_ref.data.shape:
sct.printv(
parser.error(
'ERROR: -omc inputs need to have all the same shapes'))
del im
im_out = [multicomponent_merge(fname_in)] # TODO: adapt to fname_in
elif arguments.pad is not None:
im_in = Image(fname_in[0])
ndims = len(im_in.data.shape)
if ndims != 3:
sct.printv('ERROR: you need to specify a 3D input file.', 1,
'error')
return
pad_arguments = arguments.pad.split(',')
if len(pad_arguments) != 3:
sct.printv('ERROR: you need to specify 3 padding values.', 1,
'error')
padx, pady, padz = pad_arguments
padx, pady, padz = int(padx), int(pady), int(padz)
im_out = [
pad_image(im_in,
pad_x_i=padx,
pad_x_f=padx,
pad_y_i=pady,
pad_y_f=pady,
pad_z_i=padz,
pad_z_f=padz)
]
elif arguments.pad_asym is not None:
im_in = Image(fname_in[0])
ndims = len(im_in.data.shape)
if ndims != 3:
sct.printv('ERROR: you need to specify a 3D input file.', 1,
'error')
return
pad_arguments = arguments.pad_asym.split(',')
if len(pad_arguments) != 6:
sct.printv('ERROR: you need to specify 6 padding values.', 1,
'error')
padxi, padxf, padyi, padyf, padzi, padzf = pad_arguments
padxi, padxf, padyi, padyf, padzi, padzf = int(padxi), int(padxf), int(
padyi), int(padyf), int(padzi), int(padzf)
im_out = [
pad_image(im_in,
pad_x_i=padxi,
pad_x_f=padxf,
pad_y_i=padyi,
pad_y_f=padyf,
pad_z_i=padzi,
pad_z_f=padzf)
]
elif arguments.remove_vol is not None:
index_vol = (arguments.remove_vol).split(',')
for iindex_vol, vol in enumerate(index_vol):
index_vol[iindex_vol] = int(vol)
im_in = Image(fname_in[0])
im_out = [remove_vol(im_in, index_vol, todo='remove')]
elif arguments.setorient is not None:
sct.printv(fname_in[0])
im_in = Image(fname_in[0])
im_out = [msct_image.change_orientation(im_in, arguments.setorient)]
elif arguments.setorient_data is not None:
im_in = Image(fname_in[0])
im_out = [
msct_image.change_orientation(im_in,
arguments.setorient_data,
data_only=True)
]
elif arguments.split is not None:
dim = arguments.split
assert dim in dim_list
im_in = Image(fname_in[0])
dim = dim_list.index(dim)
im_out = split_data(im_in, dim)
elif arguments.type is not None:
output_type = arguments.type
im_in = Image(fname_in[0])
im_out = [im_in] # TODO: adapt to fname_in
elif arguments.to_fsl is not None:
space_files = arguments.to_fsl
if len(space_files) > 2 or len(space_files) < 1:
sct.printv(parser.error('ERROR: -to-fsl expects 1 or 2 arguments'))
return
spaces = [Image(s) for s in space_files]
if len(spaces) < 2:
spaces.append(None)
im_out = [
displacement_to_abs_fsl(Image(fname_in[0]), spaces[0], spaces[1])
]
else:
im_out = None
sct.printv(
parser.error(
'ERROR: you need to specify an operation to do on the input image'
))
# in case fname_out is not defined, use first element of input file name list
if fname_out is None:
fname_out = fname_in[0]
# Write output
if im_out is not None:
sct.printv('Generate output files...', verbose)
# if only one output
if len(im_out) == 1 and not '-split' in arguments:
im_out[0].save(fname_out, dtype=output_type, verbose=verbose)
sct.display_viewer_syntax([fname_out], verbose=verbose)
if arguments.mcs:
# use input file name and add _X, _Y _Z. Keep the same extension
l_fname_out = []
for i_dim in range(3):
l_fname_out.append(
sct.add_suffix(fname_out or fname_in[0],
'_' + dim_list[i_dim].upper()))
im_out[i_dim].save(l_fname_out[i_dim], verbose=verbose)
sct.display_viewer_syntax(fname_out)
if arguments.split is not None:
# use input file name and add _"DIM+NUMBER". Keep the same extension
l_fname_out = []
for i, im in enumerate(im_out):
l_fname_out.append(
sct.add_suffix(
fname_out or fname_in[0],
'_' + dim_list[dim].upper() + str(i).zfill(4)))
im.save(l_fname_out[i])
sct.display_viewer_syntax(l_fname_out)
elif arguments.getorient:
sct.printv(orient)
elif arguments.display_warp:
sct.printv('Warping grid generated.', verbose, 'info')
def apply(self):
# Initialization
fname_src = self.input_filename # source image (moving)
list_warp = self.list_warp # list of warping fields
fname_out = self.output_filename # output
fname_dest = self.fname_dest # destination image (fix)
verbose = self.verbose
remove_temp_files = self.remove_temp_files
crop_reference = self.crop # if = 1, put 0 everywhere around warping field, if = 2, real crop
islabel = False
if self.interp == 'label':
islabel = True
self.interp = 'nn'
interp = get_interpolation('isct_antsApplyTransforms', self.interp)
# Parse list of warping fields
printv('\nParse list of warping fields...', verbose)
use_inverse = []
fname_warp_list_invert = []
# list_warp = list_warp.replace(' ', '') # remove spaces
# list_warp = list_warp.split(",") # parse with comma
for idx_warp, path_warp in enumerate(self.list_warp):
# Check if this transformation should be inverted
if path_warp in self.list_warpinv:
use_inverse.append('-i')
# list_warp[idx_warp] = path_warp[1:] # remove '-'
fname_warp_list_invert += [[
use_inverse[idx_warp], list_warp[idx_warp]
]]
else:
use_inverse.append('')
fname_warp_list_invert += [[path_warp]]
path_warp = list_warp[idx_warp]
if path_warp.endswith((".nii", ".nii.gz")) \
and Image(list_warp[idx_warp]).header.get_intent()[0] != 'vector':
raise ValueError(
"Displacement field in {} is invalid: should be encoded"
" in a 5D file with vector intent code"
" (see https://nifti.nimh.nih.gov/pub/dist/src/niftilib/nifti1.h"
.format(path_warp))
# need to check if last warping field is an affine transfo
isLastAffine = False
path_fname, file_fname, ext_fname = extract_fname(
fname_warp_list_invert[-1][-1])
if ext_fname in ['.txt', '.mat']:
isLastAffine = True
# check if destination file is 3d
# check_dim(fname_dest, dim_lst=[3]) # PR 2598: we decided to skip this line.
# N.B. Here we take the inverse of the warp list, because sct_WarpImageMultiTransform concatenates in the reverse order
fname_warp_list_invert.reverse()
fname_warp_list_invert = functools.reduce(lambda x, y: x + y,
fname_warp_list_invert)
# Extract path, file and extension
path_src, file_src, ext_src = extract_fname(fname_src)
path_dest, file_dest, ext_dest = extract_fname(fname_dest)
# Get output folder and file name
if fname_out == '':
path_out = '' # output in user's current directory
file_out = file_src + '_reg'
ext_out = ext_src
fname_out = os.path.join(path_out, file_out + ext_out)
# Get dimensions of data
printv('\nGet dimensions of data...', verbose)
img_src = Image(fname_src)
nx, ny, nz, nt, px, py, pz, pt = img_src.dim
# nx, ny, nz, nt, px, py, pz, pt = get_dimension(fname_src)
printv(
' ' + str(nx) + ' x ' + str(ny) + ' x ' + str(nz) + ' x ' +
str(nt), verbose)
# if 3d
if nt == 1:
# Apply transformation
printv('\nApply transformation...', verbose)
if nz in [0, 1]:
dim = '2'
else:
dim = '3'
# if labels, dilate before resampling
if islabel:
printv("\nDilate labels before warping...")
path_tmp = tmp_create(basename="apply_transfo")
fname_dilated_labels = os.path.join(path_tmp,
"dilated_data.nii")
# dilate points
dilate(Image(fname_src), 4, 'ball').save(fname_dilated_labels)
fname_src = fname_dilated_labels
printv(
"\nApply transformation and resample to destination space...",
verbose)
run_proc([
'isct_antsApplyTransforms', '-d', dim, '-i', fname_src, '-o',
fname_out, '-t'
] + fname_warp_list_invert + ['-r', fname_dest] + interp,
is_sct_binary=True)
# if 4d, loop across the T dimension
else:
if islabel:
raise NotImplementedError
dim = '4'
path_tmp = tmp_create(basename="apply_transfo")
# convert to nifti into temp folder
printv('\nCopying input data to tmp folder and convert to nii...',
verbose)
img_src.save(os.path.join(path_tmp, "data.nii"))
copy(fname_dest, os.path.join(path_tmp, file_dest + ext_dest))
fname_warp_list_tmp = []
for fname_warp in list_warp:
path_warp, file_warp, ext_warp = extract_fname(fname_warp)
copy(fname_warp, os.path.join(path_tmp, file_warp + ext_warp))
fname_warp_list_tmp.append(file_warp + ext_warp)
fname_warp_list_invert_tmp = fname_warp_list_tmp[::-1]
curdir = os.getcwd()
os.chdir(path_tmp)
# split along T dimension
printv('\nSplit along T dimension...', verbose)
im_dat = Image('data.nii')
im_header = im_dat.hdr
data_split_list = sct_image.split_data(im_dat, 3)
for im in data_split_list:
im.save()
# apply transfo
printv('\nApply transformation to each 3D volume...', verbose)
for it in range(nt):
file_data_split = 'data_T' + str(it).zfill(4) + '.nii'
file_data_split_reg = 'data_reg_T' + str(it).zfill(4) + '.nii'
status, output = run_proc([
'isct_antsApplyTransforms',
'-d',
'3',
'-i',
file_data_split,
'-o',
file_data_split_reg,
'-t',
] + fname_warp_list_invert_tmp + [
'-r',
file_dest + ext_dest,
] + interp,
verbose,
is_sct_binary=True)
# Merge files back
printv('\nMerge file back...', verbose)
import glob
path_out, name_out, ext_out = extract_fname(fname_out)
# im_list = [Image(file_name) for file_name in glob.glob('data_reg_T*.nii')]
# concat_data use to take a list of image in input, now takes a list of file names to open the files one by one (see issue #715)
fname_list = glob.glob('data_reg_T*.nii')
fname_list.sort()
im_list = [Image(fname) for fname in fname_list]
im_out = sct_image.concat_data(im_list, 3, im_header['pixdim'])
im_out.save(name_out + ext_out)
os.chdir(curdir)
generate_output_file(os.path.join(path_tmp, name_out + ext_out),
fname_out)
# Delete temporary folder if specified
if remove_temp_files:
printv('\nRemove temporary files...', verbose)
rmtree(path_tmp, verbose=verbose)
# Copy affine matrix from destination space to make sure qform/sform are the same
printv(
"Copy affine matrix from destination space to make sure qform/sform are the same.",
verbose)
im_src_reg = Image(fname_out)
im_src_reg.copy_qform_from_ref(Image(fname_dest))
im_src_reg.save(
verbose=0
) # set verbose=0 to avoid warning message about rewriting file
if islabel:
printv(
"\nTake the center of mass of each registered dilated labels..."
)
labeled_img = cubic_to_point(im_src_reg)
labeled_img.save(path=fname_out)
if remove_temp_files:
printv('\nRemove temporary files...', verbose)
rmtree(path_tmp, verbose=verbose)
# Crop the resulting image using dimensions from the warping field
warping_field = fname_warp_list_invert[-1]
# If the last transformation is not an affine transfo, we need to compute the matrix space of the concatenated
# warping field
if not isLastAffine and crop_reference in [1, 2]:
printv('Last transformation is not affine.')
if crop_reference in [1, 2]:
# Extract only the first ndim of the warping field
img_warp = Image(warping_field)
if dim == '2':
img_warp_ndim = Image(img_src.data[:, :], hdr=img_warp.hdr)
elif dim in ['3', '4']:
img_warp_ndim = Image(img_src.data[:, :, :],
hdr=img_warp.hdr)
# Set zero to everything outside the warping field
cropper = ImageCropper(Image(fname_out))
cropper.get_bbox_from_ref(img_warp_ndim)
if crop_reference == 1:
printv(
'Cropping strategy is: keep same matrix size, put 0 everywhere around warping field'
)
img_out = cropper.crop(background=0)
elif crop_reference == 2:
printv(
'Cropping strategy is: crop around warping field (the size of warping field will '
'change)')
img_out = cropper.crop()
img_out.save(fname_out)
display_viewer_syntax([fname_dest, fname_out], verbose=verbose)
def __init__(self, fname_mask, fname_sc, fname_ref, path_template,
path_ofolder, verbose):
self.fname_mask = fname_mask
self.fname_sc = fname_sc
self.fname_ref = fname_ref
self.path_template = path_template
self.path_ofolder = path_ofolder
self.verbose = verbose
self.wrk_dir = os.getcwd()
if not set(np.unique(Image(fname_mask).data)) == set([0.0, 1.0]):
if set(np.unique(Image(fname_mask).data)) == set([0.0]):
printv('WARNING: Empty masked image', self.verbose, 'warning')
else:
printv(
"ERROR input file %s is not binary file with 0 and 1 values"
% fname_mask, 1, 'error')
# create tmp directory
self.tmp_dir = tmp_create() # path to tmp directory
# lesion file where each lesion has a different value
self.fname_label = extract_fname(
self.fname_mask)[1] + '_label' + extract_fname(self.fname_mask)[2]
# initialization of measure sheet
measure_lst = [
'label', 'volume [mm3]', 'length [mm]',
'max_equivalent_diameter [mm]'
]
if self.fname_ref is not None:
for measure in ['mean', 'std']:
measure_lst.append(measure + '_' +
extract_fname(self.fname_ref)[1])
measure_dct = {}
for column in measure_lst:
measure_dct[column] = None
self.measure_pd = pd.DataFrame(data=measure_dct,
index=range(0),
columns=measure_lst)
# orientation of the input image
self.orientation = None
# volume object
self.volumes = None
# initialization of proportion measures, related to registrated atlas
if self.path_template is not None:
self.path_atlas = os.path.join(self.path_template, "atlas")
self.path_levels = os.path.join(self.path_template, "template",
"PAM50_levels.nii.gz")
else:
self.path_atlas, self.path_levels = None, None
self.vert_lst = None
self.atlas_roi_lst = None
self.distrib_matrix_dct = {}
# output names
self.pickle_name = extract_fname(self.fname_mask)[1] + '_analyzis.pkl'
self.excel_name = extract_fname(self.fname_mask)[1] + '_analyzis.xls'
# path_tmp = sct.tmp_create()
# tmp_copy_nifti(input_file, path_tmp, 'raw.nii')
# sct.run('cp '+warping_fields_filename[0]+' '+path_tmp)
# curdir = os.getcwd()
# os.chdir(path_tmp)
sct.mkdir("images")
sct.mkdir("niftis")
while True:
try:
warping_fields[0].num_of_frames = number_of_frames
image_output_iter, iteration = next(warping_fields[0])
image_output_iter.save()
filename_warp = image_output_iter.path + image_output_iter.file_name + image_output_iter.ext
filename_output = "niftis/tmp.warped_image_" + str(
iteration - 1) + image_output_iter.ext
sct.run([
"sct_apply_transfo", "-i", input_file, "-d", reference_image,
"-w", filename_warp, "-o", filename_output
])
result = Image(filename_output).change_orientation("RPI")
toimage(result.data[int(result.data.shape[0] / 2)].squeeze(),
cmin=0.0).save('images/' +
extract_fname(filename_output)[1] + '.jpg')
filenames_output.append(filename_output)
except ValueError:
printv('\nError during warping field generation...', 1, 'error')
except StopIteration:
printv('\nFinished iterations.')
break
def main(args=None):
"""
Main function
:param args:
:return:
"""
# get parser args
if args is None:
args = None if sys.argv[1:] else ['--help']
else:
# flatten the list of input arguments because -w and -winv carry a nested list
lst = []
for line in args:
lst.append(line) if isinstance(line, str) else lst.extend(line)
args = lst
parser = get_parser()
arguments = parser.parse_args(args=args)
# Initialization
fname_warp_final = '' # concatenated transformations
fname_dest = arguments.d
fname_warp_list = arguments.w
warpinv_filename = arguments.winv
if arguments.o is not None:
fname_warp_final = arguments.o
verbose = arguments.v
sct.init_sct(log_level=verbose, update=True) # Update log level
# Parse list of warping fields
sct.printv('\nParse list of warping fields...', verbose)
use_inverse = []
fname_warp_list_invert = []
# list_warp = list_warp.replace(' ', '') # remove spaces
# list_warp = list_warp.split(",") # parse with comma
for idx_warp, path_warp in enumerate(fname_warp_list):
# Check if this transformation should be inverted
if path_warp in warpinv_filename:
use_inverse.append('-i')
# list_warp[idx_warp] = path_warp[1:] # remove '-'
fname_warp_list_invert += [[
use_inverse[idx_warp], fname_warp_list[idx_warp]
]]
else:
use_inverse.append('')
fname_warp_list_invert += [[path_warp]]
path_warp = fname_warp_list[idx_warp]
if path_warp.endswith((".nii", ".nii.gz")) \
and Image(fname_warp_list[idx_warp]).header.get_intent()[0] != 'vector':
raise ValueError("Displacement field in {} is invalid: should be encoded" \
" in a 5D file with vector intent code" \
" (see https://nifti.nimh.nih.gov/pub/dist/src/niftilib/nifti1.h" \
.format(path_warp))
# need to check if last warping field is an affine transfo
isLastAffine = False
path_fname, file_fname, ext_fname = sct.extract_fname(
fname_warp_list_invert[-1][-1])
if ext_fname in ['.txt', '.mat']:
isLastAffine = True
# check if destination file is 3d
if not sct.check_dim(fname_dest, dim_lst=[3]):
sct.printv('ERROR: Destination data must be 3d')
# Here we take the inverse of the warp list, because sct_WarpImageMultiTransform concatenates in the reverse order
fname_warp_list_invert.reverse()
fname_warp_list_invert = functools.reduce(lambda x, y: x + y,
fname_warp_list_invert)
# Check file existence
sct.printv('\nCheck file existence...', verbose)
sct.check_file_exist(fname_dest, verbose)
for i in range(len(fname_warp_list)):
sct.check_file_exist(fname_warp_list[i], verbose)
# Get output folder and file name
if fname_warp_final == '':
path_out, file_out, ext_out = sct.extract_fname(param.fname_warp_final)
else:
path_out, file_out, ext_out = sct.extract_fname(fname_warp_final)
# Check dimension of destination data (cf. issue #1419, #1429)
im_dest = Image(fname_dest)
if im_dest.dim[2] == 1:
dimensionality = '2'
else:
dimensionality = '3'
cmd = [
'isct_ComposeMultiTransform', dimensionality, 'warp_final' + ext_out,
'-R', fname_dest
] + fname_warp_list_invert
status, output = sct.run(cmd, verbose=verbose, is_sct_binary=True)
# check if output was generated
if not os.path.isfile('warp_final' + ext_out):
sct.printv('ERROR: Warping field was not generated.\n' + output, 1,
'error')
# Generate output files
sct.printv('\nGenerate output files...', verbose)
sct.generate_output_file('warp_final' + ext_out,
os.path.join(path_out, file_out + ext_out))
def check_and_correct_segmentation(fname_segmentation,
fname_centerline,
folder_output='',
threshold_distance=5.0,
remove_temp_files=1,
verbose=0):
"""
This function takes the outputs of isct_propseg (centerline and segmentation) and check if the centerline of the
segmentation is coherent with the centerline provided by the isct_propseg, especially on the edges (related
to issue #1074).
Args:
fname_segmentation: filename of binary segmentation
fname_centerline: filename of binary centerline
threshold_distance: threshold, in mm, beyond which centerlines are not coherent
verbose:
Returns: None
"""
sct.printv('\nCheck consistency of segmentation...', verbose)
# creating a temporary folder in which all temporary files will be placed and deleted afterwards
path_tmp = sct.tmp_create(basename="propseg", verbose=verbose)
from sct_convert import convert
convert(fname_segmentation,
os.path.join(path_tmp, "tmp.segmentation.nii.gz"),
verbose=0)
convert(fname_centerline,
os.path.join(path_tmp, "tmp.centerline.nii.gz"),
verbose=0)
fname_seg_absolute = os.path.abspath(fname_segmentation)
# go to tmp folder
curdir = os.getcwd()
os.chdir(path_tmp)
# convert segmentation image to RPI
im_input = Image('tmp.segmentation.nii.gz')
image_input_orientation = im_input.orientation
sct_image.main(
"-i tmp.segmentation.nii.gz -setorient RPI -o tmp.segmentation_RPI.nii.gz -v 0"
.split())
sct_image.main(
"-i tmp.centerline.nii.gz -setorient RPI -o tmp.centerline_RPI.nii.gz -v 0"
.split())
# go through segmentation image, and compare with centerline from propseg
im_seg = Image('tmp.segmentation_RPI.nii.gz')
im_centerline = Image('tmp.centerline_RPI.nii.gz')
# Get size of data
sct.printv('\nGet data dimensions...', verbose)
nx, ny, nz, nt, px, py, pz, pt = im_seg.dim
# extraction of centerline provided by isct_propseg and computation of center of mass for each slice
# the centerline is defined as the center of the tubular mesh outputed by propseg.
centerline, key_centerline = {}, []
for i in range(nz):
slice = im_centerline.data[:, :, i]
if np.any(slice):
x_centerline, y_centerline = ndi.measurements.center_of_mass(slice)
centerline[str(i)] = [x_centerline, y_centerline]
key_centerline.append(i)
minz_centerline = np.min(key_centerline)
maxz_centerline = np.max(key_centerline)
mid_slice = int((maxz_centerline - minz_centerline) / 2)
# for each slice of the segmentation, check if only one object is present. If not, remove the slice from segmentation.
# If only one object (the spinal cord) is present in the slice, check if its center of mass is close to the centerline of isct_propseg.
slices_to_remove = [
False
] * nz # flag that decides if the slice must be removed
for i in range(minz_centerline, maxz_centerline + 1):
# extraction of slice
slice = im_seg.data[:, :, i]
distance = -1
label_objects, nb_labels = ndi.label(
slice) # count binary objects in the slice
if nb_labels > 1: # if there is more that one object in the slice, the slice is removed from the segmentation
slices_to_remove[i] = True
elif nb_labels == 1: # check if the centerline is coherent with the one from isct_propseg
x_centerline, y_centerline = ndi.measurements.center_of_mass(slice)
slice_nearest_coord = min(key_centerline, key=lambda x: abs(x - i))
coord_nearest_coord = centerline[str(slice_nearest_coord)]
distance = np.sqrt((
(x_centerline - coord_nearest_coord[0]) * px)**2 + (
(y_centerline - coord_nearest_coord[1]) * py)**2 +
((i - slice_nearest_coord) * pz)**2)
if distance >= threshold_distance: # threshold must be adjusted, default is 5 mm
slices_to_remove[i] = True
# Check list of removal and keep one continuous centerline (improve this comment)
# Method:
# starting from mid-centerline (in both directions), the first True encountered is applied to all following slices
slice_to_change = False
for i in range(mid_slice, nz):
if slice_to_change:
slices_to_remove[i] = True
elif slices_to_remove[i]:
slice_to_change = True
slice_to_change = False
for i in range(mid_slice, 0, -1):
if slice_to_change:
slices_to_remove[i] = True
elif slices_to_remove[i]:
slice_to_change = True
for i in range(0, nz):
# remove the slice
if slices_to_remove[i]:
im_seg.data[:, :, i] *= 0
# saving the image
im_seg.save('tmp.segmentation_RPI_c.nii.gz')
# replacing old segmentation with the corrected one
sct_image.main(
'-i tmp.segmentation_RPI_c.nii.gz -setorient {} -o {} -v 0'.format(
image_input_orientation, fname_seg_absolute).split())
os.chdir(curdir)
# display information about how much of the segmentation has been corrected
# remove temporary files
if remove_temp_files:
# sct.printv("\nRemove temporary files...", verbose)
sct.rmtree(path_tmp)
def propseg(img_input, options_dict):
"""
:param img_input: source image, to be segmented
:param options_dict: arguments as dictionary
:return: segmented Image
"""
arguments = options_dict
fname_input_data = img_input.absolutepath
fname_data = os.path.abspath(fname_input_data)
contrast_type = arguments["-c"]
contrast_type_conversion = {
't1': 't1',
't2': 't2',
't2s': 't2',
'dwi': 't1'
}
contrast_type_propseg = contrast_type_conversion[contrast_type]
# Starting building the command
cmd = ['isct_propseg', '-t', contrast_type_propseg]
if "-ofolder" in arguments:
folder_output = arguments["-ofolder"]
else:
folder_output = './'
cmd += ['-o', folder_output]
if not os.path.isdir(folder_output) and os.path.exists(folder_output):
sct.log.error("output directory %s is not a valid directory" %
folder_output)
if not os.path.exists(folder_output):
os.makedirs(folder_output)
if "-down" in arguments:
cmd += ["-down", str(arguments["-down"])]
if "-up" in arguments:
cmd += ["-up", str(arguments["-up"])]
remove_temp_files = 1
if "-r" in arguments:
remove_temp_files = int(arguments["-r"])
verbose = 0
if "-v" in arguments:
if arguments["-v"] is "1":
verbose = 2
cmd += ["-verbose"]
# Output options
if "-mesh" in arguments:
cmd += ["-mesh"]
if "-centerline-binary" in arguments:
cmd += ["-centerline-binary"]
if "-CSF" in arguments:
cmd += ["-CSF"]
if "-centerline-coord" in arguments:
cmd += ["-centerline-coord"]
if "-cross" in arguments:
cmd += ["-cross"]
if "-init-tube" in arguments:
cmd += ["-init-tube"]
if "-low-resolution-mesh" in arguments:
cmd += ["-low-resolution-mesh"]
if "-detect-nii" in arguments:
cmd += ["-detect-nii"]
if "-detect-png" in arguments:
cmd += ["-detect-png"]
# Helping options
use_viewer = None
use_optic = True # enabled by default
init_option = None
rescale_header = arguments["-rescale"]
if "-init" in arguments:
init_option = float(arguments["-init"])
if init_option < 0:
sct.log.error('Command-line usage error: ' + str(init_option) +
" is not a valid value for '-init'")
sys.exit(1)
if "-init-centerline" in arguments:
if str(arguments["-init-centerline"]) == "viewer":
use_viewer = "centerline"
elif str(arguments["-init-centerline"]) == "hough":
use_optic = False
else:
if rescale_header is not 1:
fname_labels_viewer = func_rescale_header(str(
arguments["-init-centerline"]),
rescale_header,
verbose=verbose)
else:
fname_labels_viewer = str(arguments["-init-centerline"])
cmd += ["-init-centerline", fname_labels_viewer]
use_optic = False
if "-init-mask" in arguments:
if str(arguments["-init-mask"]) == "viewer":
use_viewer = "mask"
else:
if rescale_header is not 1:
fname_labels_viewer = func_rescale_header(
str(arguments["-init-mask"]), rescale_header)
else:
fname_labels_viewer = str(arguments["-init-mask"])
cmd += ["-init-mask", fname_labels_viewer]
use_optic = False
if "-mask-correction" in arguments:
cmd += ["-mask-correction", str(arguments["-mask-correction"])]
if "-radius" in arguments:
cmd += ["-radius", str(arguments["-radius"])]
if "-detect-n" in arguments:
cmd += ["-detect-n", str(arguments["-detect-n"])]
if "-detect-gap" in arguments:
cmd += ["-detect-gap", str(arguments["-detect-gap"])]
if "-init-validation" in arguments:
cmd += ["-init-validation"]
if "-nbiter" in arguments:
cmd += ["-nbiter", str(arguments["-nbiter"])]
if "-max-area" in arguments:
cmd += ["-max-area", str(arguments["-max-area"])]
if "-max-deformation" in arguments:
cmd += ["-max-deformation", str(arguments["-max-deformation"])]
if "-min-contrast" in arguments:
cmd += ["-min-contrast", str(arguments["-min-contrast"])]
if "-d" in arguments:
cmd += ["-d", str(arguments["-d"])]
if "-distance-search" in arguments:
cmd += ["-dsearch", str(arguments["-distance-search"])]
if "-alpha" in arguments:
cmd += ["-alpha", str(arguments["-alpha"])]
# check if input image is in 3D. Otherwise itk image reader will cut the 4D image in 3D volumes and only take the first one.
image_input = Image(fname_data)
nx, ny, nz, nt, px, py, pz, pt = image_input.dim
if nt > 1:
sct.log.error(
'ERROR: your input image needs to be 3D in order to be segmented.')
path_data, file_data, ext_data = sct.extract_fname(fname_data)
# rescale header (see issue #1406)
if rescale_header is not 1:
fname_data_propseg = func_rescale_header(fname_data, rescale_header)
else:
fname_data_propseg = fname_data
# add to command
cmd += ['-i', fname_data_propseg]
# if centerline or mask is asked using viewer
if use_viewer:
from spinalcordtoolbox.gui.base import AnatomicalParams
from spinalcordtoolbox.gui.centerline import launch_centerline_dialog
params = AnatomicalParams()
if use_viewer == 'mask':
params.num_points = 3
params.interval_in_mm = 15 # superior-inferior interval between two consecutive labels
params.starting_slice = 'midfovminusinterval'
if use_viewer == 'centerline':
# setting maximum number of points to a reasonable value
params.num_points = 20
params.interval_in_mm = 30
params.starting_slice = 'top'
im_data = Image(fname_data_propseg)
im_mask_viewer = msct_image.zeros_like(im_data)
# im_mask_viewer.absolutepath = sct.add_suffix(fname_data_propseg, '_labels_viewer')
controller = launch_centerline_dialog(im_data, im_mask_viewer, params)
fname_labels_viewer = sct.add_suffix(fname_data_propseg,
'_labels_viewer')
if not controller.saved:
sct.log.error(
'The viewer has been closed before entering all manual points. Please try again.'
)
sys.exit(1)
# save labels
controller.as_niftii(fname_labels_viewer)
# add mask filename to parameters string
if use_viewer == "centerline":
cmd += ["-init-centerline", fname_labels_viewer]
elif use_viewer == "mask":
cmd += ["-init-mask", fname_labels_viewer]
# If using OptiC
elif use_optic:
path_script = os.path.dirname(__file__)
path_sct = os.path.dirname(path_script)
path_classifier = os.path.join(path_sct, 'data/optic_models',
'{}_model'.format(contrast_type))
init_option_optic, fname_centerline = optic.detect_centerline(
fname_data_propseg,
contrast_type,
path_classifier,
folder_output,
remove_temp_files,
init_option,
verbose=verbose)
if init_option is not None:
# TODO: what's this???
cmd += ["-init", str(init_option_optic)]
cmd += ["-init-centerline", fname_centerline]
# enabling centerline extraction by default (needed by check_and_correct_segmentation() )
cmd += ['-centerline-binary']
# run propseg
status, output = sct.run(cmd, verbose, raise_exception=False)
# check status is not 0
if not status == 0:
sct.log.error(
'Automatic cord detection failed. Please initialize using -init-centerline or '
'-init-mask (see help).')
sys.exit(1)
# build output filename
fname_seg = os.path.join(
folder_output, os.path.basename(sct.add_suffix(fname_data, "_seg")))
fname_centerline = os.path.join(
folder_output,
os.path.basename(sct.add_suffix(fname_data, "_centerline")))
# in case header was rescaled, we need to update the output file names by removing the "_rescaled"
if rescale_header is not 1:
sct.mv(
os.path.join(
folder_output,
sct.add_suffix(os.path.basename(fname_data_propseg), "_seg")),
fname_seg)
sct.mv(
os.path.join(
folder_output,
sct.add_suffix(os.path.basename(fname_data_propseg),
"_centerline")), fname_centerline)
# if user was used, copy the labelled points to the output folder (they will then be scaled back)
if use_viewer:
fname_labels_viewer_new = os.path.join(
folder_output,
os.path.basename(sct.add_suffix(fname_data, "_labels_viewer")))
sct.copy(fname_labels_viewer, fname_labels_viewer_new)
# update variable (used later)
fname_labels_viewer = fname_labels_viewer_new
# check consistency of segmentation
if arguments["-correct-seg"] == "1":
check_and_correct_segmentation(fname_seg,
fname_centerline,
folder_output=folder_output,
threshold_distance=3.0,
remove_temp_files=remove_temp_files,
verbose=verbose)
# copy header from input to segmentation to make sure qform is the same
sct.printv(
"Copy header input --> output(s) to make sure qform is the same.",
verbose)
list_fname = [fname_seg, fname_centerline]
if use_viewer:
list_fname.append(fname_labels_viewer)
for fname in list_fname:
im = Image(fname)
im.header = image_input.header
im.save(dtype='int8'
) # they are all binary masks hence fine to save as int8
return Image(fname_seg)
def main():
"""Main function."""
sct.init_sct()
parser = get_parser()
args = sys.argv[1:]
arguments = parser.parse(args)
fname_image = os.path.abspath(arguments['-i'])
contrast_type = arguments['-c']
ctr_algo = arguments["-centerline"]
if "-brain" not in args:
if contrast_type in ['t2s', 'dwi']:
brain_bool = False
if contrast_type in ['t1', 't2']:
brain_bool = True
else:
brain_bool = bool(int(arguments["-brain"]))
kernel_size = arguments["-kernel"]
if kernel_size == '3d' and contrast_type == 'dwi':
kernel_size = '2d'
sct.printv('3D kernel model for dwi contrast is not available. 2D kernel model is used instead.', type="warning")
if '-ofolder' not in args:
output_folder = os.getcwd()
else:
output_folder = arguments["-ofolder"]
if ctr_algo == 'file' and "-file_centerline" not in args:
sct.log.warning('Please use the flag -file_centerline to indicate the centerline filename.')
sys.exit(1)
if "-file_centerline" in args:
manual_centerline_fname = arguments["-file_centerline"]
ctr_algo = 'file'
else:
manual_centerline_fname = None
remove_temp_files = int(arguments['-r'])
verbose = int(arguments['-v'])
path_qc = arguments.get("-qc", None)
algo_config_stg = '\nMethod:'
algo_config_stg += '\n\tCenterline algorithm: ' + str(ctr_algo)
algo_config_stg += '\n\tAssumes brain section included in the image: ' + str(brain_bool)
algo_config_stg += '\n\tDimension of the segmentation kernel convolutions: ' + kernel_size + '\n'
sct.printv(algo_config_stg)
im_image = Image(fname_image)
# note: below we pass im_image.copy() otherwise the field absolutepath becomes None after execution of this function
im_seg, im_image_RPI_upsamp, im_seg_RPI_upsamp, im_labels_viewer, im_ctr = deep_segmentation_spinalcord(
im_image.copy(), contrast_type, ctr_algo=ctr_algo, ctr_file=manual_centerline_fname,
brain_bool=brain_bool, kernel_size=kernel_size, remove_temp_files=remove_temp_files, verbose=verbose)
# Save segmentation
fname_seg = os.path.abspath(os.path.join(output_folder, sct.extract_fname(fname_image)[1] + '_seg' +
sct.extract_fname(fname_image)[2]))
im_seg.save(fname_seg)
if ctr_algo == 'viewer':
# Save labels
fname_labels = os.path.abspath(os.path.join(output_folder, sct.extract_fname(fname_image)[1] + '_labels-centerline' +
sct.extract_fname(fname_image)[2]))
im_labels_viewer.save(fname_labels)
if verbose == 2:
# Save ctr
fname_ctr = os.path.abspath(os.path.join(output_folder, sct.extract_fname(fname_image)[1] + '_centerline' +
sct.extract_fname(fname_image)[2]))
im_ctr.save(fname_ctr)
if path_qc is not None:
generate_qc(fname_image, fname_seg=fname_seg, args=args, path_qc=os.path.abspath(path_qc),
process='sct_deepseg_sc')
sct.display_viewer_syntax([fname_image, fname_seg], colormaps=['gray', 'red'], opacities=['', '0.7'])
def test_integrity(param_test):
"""
Test integrity of function
"""
if param_test.args.startswith(default_args[1]):
return param_test # no integrity test
# apply transformation to binary mask: template --> anat
sct_apply_transfo.main(args=[
'-i', param_test.fname_gt, '-d', param_test.dict_args_with_path['-s'],
'-w',
os.path.join(param_test.path_output,
'warp_template2anat.nii.gz'), '-o',
os.path.join(param_test.path_output,
'test_template2anat.nii.gz'), '-x', 'nn', '-v', '0'
])
# apply transformation to binary mask: anat --> template
sct_apply_transfo.main(args=[
'-i', param_test.dict_args_with_path['-s'], '-d', param_test.fname_gt,
'-w',
os.path.join(param_test.path_output,
'warp_anat2template.nii.gz'), '-o',
os.path.join(param_test.path_output,
'test_anat2template.nii.gz'), '-x', 'nn', '-v', '0'
])
# compute dice coefficient between template segmentation warped to anat and segmentation from anat
im_seg = Image(param_test.dict_args_with_path['-s'])
im_template_seg_reg = Image(
os.path.join(param_test.path_output, 'test_template2anat.nii.gz'))
dice_template2anat = msct_image.compute_dice(im_seg,
im_template_seg_reg,
mode='3d',
zboundaries=True)
# check
param_test.output += 'Dice[seg,template_seg_reg]: ' + str(
dice_template2anat)
if dice_template2anat > param_test.dice_threshold:
param_test.output += '\n--> PASSED'
else:
param_test.status = 99
param_test.output += '\n--> FAILED'
# compute dice coefficient between anat segmentation warped to template and segmentation from template
im_seg_reg = Image(
os.path.join(param_test.path_output, 'test_anat2template.nii.gz'))
im_template_seg = Image(param_test.fname_gt)
dice_anat2template = msct_image.compute_dice(im_seg_reg,
im_template_seg,
mode='3d',
zboundaries=True)
# check
param_test.output += '\n\nDice[seg_reg,template_seg]: ' + str(
dice_anat2template)
if dice_anat2template > param_test.dice_threshold:
param_test.output += '\n--> PASSED'
else:
param_test.status = 99
param_test.output += '\n--> FAILED'
# update Panda structure
param_test.results['dice_template2anat'] = dice_template2anat
param_test.results['dice_anat2template'] = dice_anat2template
return param_test
def main(fname_data, path_label, method, slices, levels, fname_output, labels_user, append_csv,
fname_vertebral_labeling="", perslice=1, perlevel=1, verbose=1, combine_labels=True):
"""
Extract metrics from MRI data based on mask (could be single file of folder to atlas)
:param fname_data: data to extract metric from
:param path_label: mask: could be single file or folder to atlas (which contains info_label.txt)
:param method {'wa', 'bin', 'ml', 'map'}
:param slices. Slices of interest. Accepted format:
"0,1,2,3": slices 0,1,2,3
"0:3": slices 0,1,2,3
:param levels: Vertebral levels to extract metrics from. Should be associated with a template
(e.g. PAM50/template/) or a specified file: fname_vertebral_labeling. Same format as slices_of_interest.
:param fname_output:
:param labels_user:
:param append_csv: Append to csv file
:param fname_normalizing_label:
:param fname_vertebral_labeling: vertebral labeling to be used with vertebral_levels
:param perslice: if user selected several slices, then the function outputs a metric within each slice
instead of a single average output.
:param perlevel: if user selected several levels, then the function outputs a metric within each vertebral level
instead of a single average output.
:param verbose
:param combine_labels: bool: Combine labels into a single value
:return:
"""
# check if path_label is a file (e.g., single binary mask) instead of a folder (e.g., SCT atlas structure which
# contains info_label.txt file)
if os.path.isfile(path_label):
# Label is a single file
indiv_labels_ids = [0]
indiv_labels_files = [path_label]
combined_labels_ids = []
label_struc = {0: LabelStruc(id=0,
name=sct.extract_fname(path_label)[1],
filename=path_label)}
# set path_label to empty string, because indiv_labels_files will replace it from now on
path_label = ''
elif os.path.isdir(path_label):
# Labels is an SCT atlas folder structure
# Parse labels according to the file info_label.txt
# Note: the "combined_labels_*" is a list of single labels that are defined in the section defined by the keyword
# "# Keyword=CombinedLabels" in info_label.txt.
# TODO: redirect to appropriate Sphinx documentation
# TODO: output Class instead of multiple variables.
# Example 1:
# label_struc[2].id = (2)
# label_struc[2].name = "left fasciculus cuneatus"
# label_struc[2].filename = "PAM50_atlas_02.nii.gz"
# Example 2:
# label_struc[51].id = (1, 2, 3, ..., 29)
# label_struc[51].name = "White Matter"
# label_struc[51].filename = "" # no name because it is combined
indiv_labels_ids, indiv_labels_names, indiv_labels_files, \
combined_labels_ids, combined_labels_names, combined_labels_id_groups, map_clusters \
= read_label_file(path_label, param_default.file_info_label)
label_struc = {}
# fill IDs for indiv labels
for i_label in range(len(indiv_labels_ids)):
label_struc[indiv_labels_ids[i_label]] = LabelStruc(id=indiv_labels_ids[i_label],
name=indiv_labels_names[i_label],
filename=indiv_labels_files[i_label],
map_cluster=[indiv_labels_ids[i_label] in map_cluster for
map_cluster in map_clusters].index(True))
# fill IDs for combined labels
# TODO: problem for defining map_cluster: if labels overlap two regions, e.g. WM and GM (e.g. id=50),
# map_cluster will take value 0, which is wrong.
for i_label in range(len(combined_labels_ids)):
label_struc[combined_labels_ids[i_label]] = LabelStruc(id=combined_labels_id_groups[i_label],
name=combined_labels_names[i_label],
map_cluster=[indiv_labels_ids[i_label] in map_cluster for
map_cluster in map_clusters].index(True))
else:
raise RuntimeError(path_label + ' does not exist')
# check syntax of labels asked by user
labels_id_user = check_labels(indiv_labels_ids + combined_labels_ids, parse_num_list(labels_user))
nb_labels = len(indiv_labels_files)
# Load data and systematically reorient to RPI because we need the 3rd dimension to be z
sct.printv('\nLoad metric image...', verbose)
input_im = Image(fname_data).change_orientation("RPI")
data = Metric(data=input_im.data, label='')
# Load labels
labels_tmp = np.empty([nb_labels], dtype=object)
for i_label in range(nb_labels):
im_label = Image(os.path.join(path_label, indiv_labels_files[i_label])).change_orientation("RPI")
labels_tmp[i_label] = np.expand_dims(im_label.data, 3) # TODO: generalize to 2D input label
labels = np.concatenate(labels_tmp[:], 3) # labels: (x,y,z,label)
# Load vertebral levels
if vertebral_levels:
im_vertebral_labeling = Image(fname_vertebral_labeling).change_orientation("RPI")
else:
im_vertebral_labeling = None
# Get dimensions of data and labels
nx, ny, nz = data.data.shape
nx_atlas, ny_atlas, nz_atlas, nt_atlas = labels.shape
# Check dimensions consistency between atlas and data
if (nx, ny, nz) != (nx_atlas, ny_atlas, nz_atlas):
sct.printv('\nERROR: Metric data and labels DO NOT HAVE SAME DIMENSIONS.', 1, type='error')
# Combine individual labels for estimation
if combine_labels:
# Add entry with internal ID value (99) which corresponds to combined labels
label_struc[99] = LabelStruc(id=labels_id_user, name=','.join([str(i) for i in labels_id_user]),
map_cluster=None)
labels_id_user = [99]
for id_label in labels_id_user:
sct.printv('Estimation for label: '+label_struc[id_label].name, verbose)
agg_metric = extract_metric(data, labels=labels, slices=slices, levels=levels, perslice=perslice,
perlevel=perlevel, vert_level=im_vertebral_labeling, method=method,
label_struc=label_struc, id_label=id_label, indiv_labels_ids=indiv_labels_ids)
save_as_csv(agg_metric, fname_output, fname_in=fname_data, append=append_csv)
append_csv = True # when looping across labels, need to append results in the same file
sct.display_open(fname_output)
def main():
# Default params
param = Param()
# Get parser info
parser = get_parser()
arguments = parser.parse(sys.argv[1:])
fname_data = arguments['-i']
if '-m' in arguments:
fname_mask = arguments['-m']
else:
fname_mask = ''
method = arguments["-method"]
if '-vol' in arguments:
index_vol_user = arguments['-vol']
else:
index_vol_user = ''
# Check parameters
if method == 'diff':
if not fname_mask:
sct.printv('You need to provide a mask with -method diff. Exit.',
1,
type='error')
# Load data and orient to RPI
im_data = Image(fname_data).change_orientation('RPI')
data = im_data.data
if fname_mask:
mask = Image(fname_mask).change_orientation('RPI').data
# Retrieve selected volumes
if index_vol_user:
index_vol = parse_num_list(index_vol_user)
else:
index_vol = range(data.shape[3])
# Make sure user selected 2 volumes with diff method
if method == 'diff':
if not len(index_vol) == 2:
sct.printv(
'Method "diff" should be used with exactly two volumes (specify with flag "-vol").',
1, 'error')
# Compute SNR
# NB: "time" is assumed to be the 4th dimension of the variable "data"
if method == 'mult':
# Compute mean and STD across time
data_mean = np.mean(data[:, :, :, index_vol], axis=3)
data_std = np.std(data[:, :, :, index_vol], axis=3)
# Generate mask where std is different from 0
mask_std_nonzero = np.where(data_std > param.almost_zero)
snr_map = np.zeros_like(data_mean)
snr_map[mask_std_nonzero] = data_mean[mask_std_nonzero] / data_std[
mask_std_nonzero]
# Output SNR map
fname_snr = sct.add_suffix(fname_data, '_SNR-' + method)
im_snr = empty_like(im_data)
im_snr.data = snr_map
im_snr.save(fname_snr, dtype=np.float32)
# Output non-zero mask
fname_stdnonzero = sct.add_suffix(fname_data,
'_mask-STD-nonzero' + method)
im_stdnonzero = empty_like(im_data)
data_stdnonzero = np.zeros_like(data_mean)
data_stdnonzero[mask_std_nonzero] = 1
im_stdnonzero.data = data_stdnonzero
im_stdnonzero.save(fname_stdnonzero, dtype=np.float32)
# Compute SNR in ROI
if fname_mask:
mean_in_roi = np.average(data_mean[mask_std_nonzero],
weights=mask[mask_std_nonzero])
std_in_roi = np.average(data_std[mask_std_nonzero],
weights=mask[mask_std_nonzero])
snr_roi = mean_in_roi / std_in_roi
# snr_roi = np.average(snr_map[mask_std_nonzero], weights=mask[mask_std_nonzero])
elif method == 'diff':
data_2vol = np.take(data, index_vol, axis=3)
# Compute mean in ROI
data_mean = np.mean(data_2vol, axis=3)
mean_in_roi = np.average(data_mean, weights=mask)
data_sub = np.subtract(data_2vol[:, :, :, 1], data_2vol[:, :, :, 0])
_, std_in_roi = weighted_avg_and_std(data_sub, mask)
# Compute SNR, correcting for Rayleigh noise (see eq. 7 in Dietrich et al.)
snr_roi = (2 / np.sqrt(2)) * mean_in_roi / std_in_roi
# Display result
if fname_mask:
sct.printv('\nSNR_' + method + ' = ' + str(snr_roi) + '\n',
type='info')
def resample_nib(image, new_size=None, new_size_type=None, image_dest=None, interpolation='linear', mode='nearest'):
"""
Resample a nibabel or Image object based on a specified resampling factor.
Can deal with 2d, 3d or 4d image objects.
:param image: nibabel or Image image.
:param new_size: list of float: Resampling factor, final dimension or resolution, depending on new_size_type.
:param new_size_type: {'vox', 'factor', 'mm'}: Feature used for resampling. Examples:
new_size=[128, 128, 90], new_size_type='vox' --> Resampling to a dimension of 128x128x90 voxels
new_size=[2, 2, 2], new_size_type='factor' --> 2x isotropic upsampling
new_size=[1, 1, 5], new_size_type='mm' --> Resampling to a resolution of 1x1x5 mm
:param image_dest: Destination image to resample the input image to. In this case, new_size and new_size_type
are ignored
:param interpolation: {'nn', 'linear', 'spline'}. The interpolation type
:param mode: Outside values are filled with 0 ('constant') or nearest value ('nearest').
:return: The resampled nibabel or Image image (depending on the input object type).
"""
# set interpolation method
dict_interp = {'nn': 0, 'linear': 1, 'spline': 2}
# If input is an Image object, create nibabel object from it
if type(image) == nib.nifti1.Nifti1Image:
img = image
elif type(image) == Image:
img = nib.nifti1.Nifti1Image(image.data, image.hdr.get_best_affine())
else:
raise Exception(TypeError)
if image_dest is None:
# Get dimensions of data
p = img.header.get_zooms()
shape = img.header.get_data_shape()
if img.ndim == 4:
new_size += ['1'] # needed because the code below is general, i.e., does not assume 3d input and uses img.shape
# compute new shape based on specific resampling method
if new_size_type == 'vox':
shape_r = tuple([int(new_size[i]) for i in range(img.ndim)])
elif new_size_type == 'factor':
if len(new_size) == 1:
# isotropic resampling
new_size = tuple([new_size[0] for i in range(img.ndim)])
# compute new shape as: shape_r = shape * f
shape_r = tuple([int(np.round(shape[i] * float(new_size[i]))) for i in range(img.ndim)])
elif new_size_type == 'mm':
if len(new_size) == 1:
# isotropic resampling
new_size = tuple([new_size[0] for i in range(img.ndim)])
# compute new shape as: shape_r = shape * (p_r / p)
shape_r = tuple([int(np.round(shape[i] * float(p[i]) / float(new_size[i]))) for i in range(img.ndim)])
else:
logger.error('new_size_type is not recognized.')
# Generate 3d affine transformation: R
affine = img.affine[:4, :4]
affine[3, :] = np.array([0, 0, 0, 1]) # satisfy to nifti convention. Otherwise it grabs the temporal
logger.debug('Affine matrix: \n' + str(affine))
R = np.eye(4)
for i in range(3):
R[i, i] = img.shape[i] / float(shape_r[i])
affine_r = np.dot(affine, R)
reference = (shape_r, affine_r)
# If reference is provided
else:
if type(image_dest) == nib.nifti1.Nifti1Image:
reference = image_dest
elif type(image_dest) == Image:
reference = nib.nifti1.Nifti1Image(image_dest.data, image_dest.hdr.get_best_affine())
else:
raise Exception(TypeError)
if img.ndim == 3:
# we use mode 'nearest' to overcome issue #2453
img_r = resample_from_to(
img, to_vox_map=reference, order=dict_interp[interpolation], mode=mode, cval=0.0, out_class=None)
elif img.ndim == 4:
# TODO: Cover img_dest with 4D volumes
# Import here instead of top of the file because this is an isolated case and nibabel takes time to import
data4d = np.zeros(shape_r)
# Loop across 4th dimension and resample each 3d volume
for it in range(img.shape[3]):
# Create dummy 3d nibabel image
nii_tmp = nib.nifti1.Nifti1Image(img.get_data()[..., it], affine)
img3d_r = resample_from_to(
nii_tmp, to_vox_map=(shape_r[:-1], affine_r), order=dict_interp[interpolation], mode=mode,
cval=0.0, out_class=None)
data4d[..., it] = img3d_r.get_data()
# Create 4d nibabel Image
img_r = nib.nifti1.Nifti1Image(data4d, affine_r)
# Convert back to proper type
if type(image) == nib.nifti1.Nifti1Image:
return img_r
elif type(image) == Image:
return Image(img_r.get_data(), hdr=img_r.header, orientation=image.orientation, dim=img_r.header.get_data_shape())
def main(argv=None):
"""Main function."""
parser = get_parser()
arguments = parser.parse_args(argv)
verbose = arguments.v
set_loglevel(verbose=verbose)
fname_image = arguments.i
contrast_type = arguments.c
ctr_algo = arguments.centerline
brain_bool = bool(arguments.brain)
if arguments.brain is None and contrast_type in ['t2s', 't2_ax']:
brain_bool = False
output_folder = arguments.ofolder
if ctr_algo == 'file' and arguments.file_centerline is None:
printv(
'Please use the flag -file_centerline to indicate the centerline filename.',
1, 'error')
sys.exit(1)
if arguments.file_centerline is not None:
manual_centerline_fname = arguments.file_centerline
ctr_algo = 'file'
else:
manual_centerline_fname = None
remove_temp_files = arguments.r
algo_config_stg = '\nMethod:'
algo_config_stg += '\n\tCenterline algorithm: ' + str(ctr_algo)
algo_config_stg += '\n\tAssumes brain section included in the image: ' + str(
brain_bool) + '\n'
printv(algo_config_stg)
# Segment image
from spinalcordtoolbox.image import Image
from spinalcordtoolbox.deepseg_lesion.core import deep_segmentation_MSlesion
im_image = Image(fname_image)
im_seg, im_labels_viewer, im_ctr = deep_segmentation_MSlesion(
im_image,
contrast_type,
ctr_algo=ctr_algo,
ctr_file=manual_centerline_fname,
brain_bool=brain_bool,
remove_temp_files=remove_temp_files,
verbose=verbose)
# Save segmentation
fname_seg = os.path.abspath(
os.path.join(
output_folder,
extract_fname(fname_image)[1] + '_lesionseg' +
extract_fname(fname_image)[2]))
im_seg.save(fname_seg)
if ctr_algo == 'viewer':
# Save labels
fname_labels = os.path.abspath(
os.path.join(
output_folder,
extract_fname(fname_image)[1] + '_labels-centerline' +
extract_fname(fname_image)[2]))
im_labels_viewer.save(fname_labels)
if verbose == 2:
# Save ctr
fname_ctr = os.path.abspath(
os.path.join(
output_folder,
extract_fname(fname_image)[1] + '_centerline' +
extract_fname(fname_image)[2]))
im_ctr.save(fname_ctr)
display_viewer_syntax([fname_image, fname_seg],
colormaps=['gray', 'red'],
opacities=['', '0.7'])
tmp_dir = sct.tmp_create()
im1_name = "im1.nii.gz"
sct.copy(input_fname, os.path.join(tmp_dir, im1_name))
if input_second_fname != '':
im2_name = 'im2.nii.gz'
sct.copy(input_second_fname, os.path.join(tmp_dir, im2_name))
else:
im2_name = None
curdir = os.getcwd()
os.chdir(tmp_dir)
# now = time.time()
input_im1 = Image(
resample_image(im1_name,
binary=True,
thr=0.5,
npx=resample_to,
npy=resample_to))
input_im1.absolutepath = os.path.basename(input_fname)
if im2_name is not None:
input_im2 = Image(
resample_image(im2_name,
binary=True,
thr=0.5,
npx=resample_to,
npy=resample_to))
input_im2.absolutepath = os.path.basename(input_second_fname)
else:
input_im2 = None
computation = ComputeDistances(input_im1, im2=input_im2, param=param)
def vertebral_detection(fname,
fname_seg,
contrast,
param,
init_disc,
verbose=1,
path_template='',
path_output='../',
scale_dist=1.):
"""
Find intervertebral discs in straightened image using template matching
:param fname: file name of straigthened spinal cord
:param fname_seg: file name of straigthened spinal cord segmentation
:param contrast: t1 or t2
:param param: advanced parameters
:param init_disc:
:param verbose:
:param path_template:
:param path_output: output path for verbose=2 pictures
:param scale_dist: float: Scaling factor to adjust average distance between two adjacent intervertebral discs
:return:
"""
sct.printv('\nLook for template...', verbose)
sct.printv('Path template: ' + path_template, verbose)
# adjust file names if MNI-Poly-AMU template is used (by default: PAM50)
fname_level = get_file_label(os.path.join(path_template, 'template'),
'vertebral labeling',
output='filewithpath')
fname_template = get_file_label(os.path.join(path_template, 'template'),
contrast.upper() + '-weighted template',
output='filewithpath')
# Open template and vertebral levels
sct.printv('\nOpen template and vertebral levels...', verbose)
data_template = Image(fname_template).data
data_disc_template = Image(fname_level).data
# open anatomical volume
im_input = Image(fname)
data = im_input.data
# smooth data
data = gaussian_filter(data,
param.smooth_factor,
output=None,
mode="reflect")
# get dimension of src
nx, ny, nz = data.shape
# define xc and yc (centered in the field of view)
xc = int(np.round(nx / 2)) # direction RL
yc = int(np.round(ny / 2)) # direction AP
# get dimension of template
nxt, nyt, nzt = data_template.shape
# define xc and yc (centered in the field of view)
xct = int(np.round(nxt / 2)) # direction RL
yct = int(np.round(nyt / 2)) # direction AP
# define mean distance (in voxel) between adjacent discs: [C1/C2 -> C2/C3], [C2/C3 -> C4/C5], ..., [L1/L2 -> L2/L3]
centerline_level = data_disc_template[xct, yct, :]
# attribute value to each disc. Starts from max level, then decrease.
# NB: value 2 means disc C2/C3 (and so on and so forth).
min_level = centerline_level[centerline_level.nonzero()].min()
max_level = centerline_level[centerline_level.nonzero()].max()
list_disc_value_template = list(range(min_level, max_level))
# add disc above top one
list_disc_value_template.insert(int(0), min_level - 1)
sct.printv('\nDisc values from template: ' + str(list_disc_value_template),
verbose)
# get diff to find transitions (i.e., discs)
diff_centerline_level = np.diff(centerline_level)
# get disc z-values
list_disc_z_template = diff_centerline_level.nonzero()[0].tolist()
list_disc_z_template.reverse()
sct.printv('Z-values for each disc: ' + str(list_disc_z_template), verbose)
list_distance_template = (
np.diff(list_disc_z_template) *
(-1)).tolist() # multiplies by -1 to get positive distances
# Update distance with scaling factor
list_distance_template = [i * scale_dist for i in list_distance_template]
sct.printv(
'Distances between discs (in voxel): ' + str(list_distance_template),
verbose)
# display init disc
if verbose == 2:
import matplotlib
matplotlib.use('Agg')
import matplotlib.pyplot as plt
# get percentile for automatic contrast adjustment
data_display = np.mean(data[xc - param.size_RL:xc +
param.size_RL, :, :],
axis=0).transpose()
percmin = np.percentile(data_display, 10)
percmax = np.percentile(data_display, 90)
# display image
plt.matshow(data_display,
fignum=50,
cmap=plt.cm.gray,
clim=[percmin, percmax],
origin='lower')
plt.title('Anatomical image')
plt.autoscale(
enable=False) # to prevent autoscale of axis when displaying plot
plt.figure(50), plt.scatter(yc + param.shift_AP_visu,
init_disc[0],
c='yellow',
s=50)
plt.text(yc + param.shift_AP_visu + 4,
init_disc[0],
str(init_disc[1]) + '/' + str(init_disc[1] + 1),
verticalalignment='center',
horizontalalignment='left',
color='pink',
fontsize=15), plt.draw()
# plt.ion() # enables interactive mode
# FIND DISCS
# ===========================================================================
sct.printv('\nDetect intervertebral discs...', verbose)
# assign initial z and disc
current_z = init_disc[0]
current_disc = init_disc[1]
# create list for z and disc
list_disc_z = []
list_disc_value = []
zrange = list(range(-10, 10))
direction = 'superior'
search_next_disc = True
while search_next_disc:
sct.printv(
'Current disc: ' + str(current_disc) + ' (z=' + str(current_z) +
'). Direction: ' + direction, verbose)
try:
# get z corresponding to current disc on template
current_z_template = list_disc_z_template[current_disc]
except:
# in case reached the bottom (see issue #849)
sct.printv(
'WARNING: Reached the bottom of the template. Stop searching.',
verbose, 'warning')
break
# find next disc
# N.B. Do not search for C1/C2 disc (because poorly visible), use template distance instead
if current_disc != 1:
current_z = compute_corr_3d(data,
data_template,
x=xc,
xshift=0,
xsize=param.size_RL,
y=yc,
yshift=param.shift_AP,
ysize=param.size_AP,
z=current_z,
zshift=0,
zsize=param.size_IS,
xtarget=xct,
ytarget=yct,
ztarget=current_z_template,
zrange=zrange,
verbose=verbose,
save_suffix='_disc' +
str(current_disc),
gaussian_std=999,
path_output=path_output)
# display new disc
if verbose == 2:
plt.figure(50), plt.scatter(yc + param.shift_AP_visu,
current_z,
c='yellow',
s=50)
plt.text(yc + param.shift_AP_visu + 4,
current_z,
str(current_disc) + '/' + str(current_disc + 1),
verticalalignment='center',
horizontalalignment='left',
color='yellow',
fontsize=15), plt.draw()
# append to main list
if direction == 'superior':
# append at the beginning
list_disc_z.insert(0, current_z)
list_disc_value.insert(0, current_disc)
elif direction == 'inferior':
# append at the end
list_disc_z.append(current_z)
list_disc_value.append(current_disc)
# adjust correcting factor based on already-identified discs
if len(list_disc_z) > 1:
# compute distance between already-identified discs
list_distance_current = (np.diff(list_disc_z) * (-1)).tolist()
# retrieve the template distance corresponding to the already-identified discs
index_disc_identified = [
i for i, j in enumerate(list_disc_value_template)
if j in list_disc_value[:-1]
]
list_distance_template_identified = [
list_distance_template[i] for i in index_disc_identified
]
# divide subject and template distances for the identified discs
list_subject_to_template_distance = [
float(list_distance_current[i]) /
list_distance_template_identified[i]
for i in range(len(list_distance_current))
]
# average across identified discs to obtain an average correcting factor
correcting_factor = np.mean(list_subject_to_template_distance)
sct.printv('.. correcting factor: ' + str(correcting_factor),
verbose)
else:
correcting_factor = 1
# update list_distance specific for the subject
list_distance = [
int(np.round(list_distance_template[i] * correcting_factor))
for i in range(len(list_distance_template))
]
# assign new current_z and disc value
if direction == 'superior':
try:
approx_distance_to_next_disc = list_distance[
list_disc_value_template.index(current_disc - 1)]
except ValueError:
sct.printv(
'WARNING: Disc value not included in template. Using previously-calculated distance: '
+ str(approx_distance_to_next_disc))
# assign new current_z and disc value
current_z = current_z + approx_distance_to_next_disc
current_disc = current_disc - 1
elif direction == 'inferior':
try:
approx_distance_to_next_disc = list_distance[
list_disc_value_template.index(current_disc)]
except:
sct.printv(
'WARNING: Disc value not included in template. Using previously-calculated distance: '
+ str(approx_distance_to_next_disc))
# assign new current_z and disc value
current_z = current_z - approx_distance_to_next_disc
current_disc = current_disc + 1
# if current_z is larger than searching zone, switch direction (and start from initial z minus approximate
# distance from updated template distance)
if current_z >= nz or current_disc == 0:
sct.printv('.. Switching to inferior direction.', verbose)
direction = 'inferior'
current_disc = init_disc[1] + 1
current_z = init_disc[0] - list_distance[
list_disc_value_template.index(current_disc)]
# if current_z is lower than searching zone, stop searching
if current_z <= 0:
search_next_disc = False
# if upper disc is not 1, add disc above top disc based on mean_distance_adjusted
upper_disc = min(list_disc_value)
# if not upper_disc == 1:
sct.printv(
'Adding top disc based on adjusted template distance: #' +
str(upper_disc - 1), verbose)
approx_distance_to_next_disc = list_distance[
list_disc_value_template.index(upper_disc - 1)]
next_z = max(list_disc_z) + approx_distance_to_next_disc
sct.printv('.. approximate distance: ' + str(approx_distance_to_next_disc),
verbose)
# make sure next disc does not go beyond FOV in superior direction
if next_z > nz:
list_disc_z.insert(0, nz)
else:
list_disc_z.insert(0, next_z)
# assign disc value
list_disc_value.insert(0, upper_disc - 1)
# Label segmentation
label_segmentation(fname_seg,
list_disc_z,
list_disc_value,
verbose=verbose)
# save figure
if verbose == 2:
plt.figure(50), plt.savefig(
os.path.join(path_output, "fig_anat_straight_with_labels.png"))
def resample_image(fname,
suffix='_resampled.nii.gz',
binary=False,
npx=0.3,
npy=0.3,
thr=0.0,
interpolation='spline'):
"""
Resampling function: add a padding, resample, crop the padding
:param fname: name of the image file to be resampled
:param suffix: suffix added to the original fname after resampling
:param binary: boolean, image is binary or not
:param npx: new pixel size in the x direction
:param npy: new pixel size in the y direction
:param thr: if the image is binary, it will be thresholded at thr (default=0) after the resampling
:param interpolation: type of interpolation used for the resampling
:return: file name after resampling (or original fname if it was already in the correct resolution)
"""
im_in = Image(fname)
orientation = im_in.orientation
if orientation != 'RPI':
fname = im_in.change_orientation(
im_in, 'RPI', generate_path=True).save().absolutepath
nx, ny, nz, nt, px, py, pz, pt = im_in.dim
if np.round(px, 2) != np.round(npx, 2) or np.round(py, 2) != np.round(
npy, 2):
name_resample = sct.extract_fname(fname)[1] + suffix
if binary:
interpolation = 'nn'
if nz == 1: # when data is 2d: we convert it to a 3d image in order to avoid nipy problem of conversion nifti-->nipy with 2d data
sct.run([
'sct_image', '-i', ','.join([fname, fname]), '-concat', 'z',
'-o', fname
])
sct.run([
'sct_resample', '-i', fname, '-mm',
str(npx) + 'x' + str(npy) + 'x' + str(pz), '-o', name_resample,
'-x', interpolation
])
if nz == 1: # when input data was 2d: re-convert data 3d-->2d
sct.run(['sct_image', '-i', name_resample, '-split', 'z'])
im_split = Image(
name_resample.split('.nii.gz')[0] + '_Z0000.nii.gz')
im_split.save(name_resample)
if binary:
sct.run([
'sct_maths', '-i', name_resample, '-bin',
str(thr), '-o', name_resample
])
if orientation != 'RPI':
name_resample = Image(name_resample) \
.change_orientation(orientation, generate_path=True) \
.save() \
.absolutepath
return name_resample
else:
if orientation != 'RPI':
fname = sct.add_suffix(fname, "_RPI")
im_in = msct_image.change_orientation(im_in,
orientation).save(fname)
sct.printv('Image resolution already ' + str(npx) + 'x' + str(npy) +
'xpz')
return fname
def measure(self):
im_lesion = Image(self.fname_label)
im_lesion_data = im_lesion.data
p_lst = im_lesion.dim[4:7] # voxel size
label_lst = [l for l in np.unique(im_lesion_data)
if l] # lesion label IDs list
if self.path_template is not None:
if os.path.isfile(self.path_levels):
img_vert = Image(self.path_levels)
im_vert_data = img_vert.data
self.vert_lst = [
v for v in np.unique(im_vert_data) if v
] # list of vertebral levels available in the input image
else:
im_vert_data = None
printv(
'ERROR: the file ' + self.path_levels +
' does not exist. Please make sure the template was correctly registered and warped (sct_register_to_template or sct_register_multimodal and sct_warp_template)',
type='error')
# In order to open atlas images only one time
atlas_data_dct = {
} # dict containing the np.array of the registrated atlas
for fname_atlas_roi in self.atlas_roi_lst:
tract_id = int(
fname_atlas_roi.split('_')[-1].split('.nii.gz')[0])
img_cur = Image(fname_atlas_roi)
img_cur_copy = img_cur.copy()
atlas_data_dct[tract_id] = img_cur_copy.data
del img_cur
self.volumes = np.zeros((im_lesion.dim[2], len(label_lst)))
# iteration across each lesion to measure statistics
for lesion_label in label_lst:
im_lesion_data_cur = np.copy(im_lesion_data == lesion_label)
printv('\nMeasures on lesion #' + str(lesion_label) + '...',
self.verbose, 'normal')
label_idx = self.measure_pd[self.measure_pd.label ==
lesion_label].index
self._measure_volume(im_lesion_data_cur, p_lst, label_idx)
self._measure_length(im_lesion_data_cur, p_lst, label_idx)
self._measure_diameter(im_lesion_data_cur, p_lst, label_idx)
# compute lesion distribution for each lesion
if self.path_template is not None:
self._measure_eachLesion_distribution(
lesion_id=lesion_label,
atlas_data=atlas_data_dct,
im_vert=im_vert_data,
im_lesion=im_lesion_data_cur,
p_lst=p_lst)
if self.path_template is not None:
# compute total lesion distribution
self._measure_totLesion_distribution(
im_lesion=np.copy(im_lesion_data > 0),
atlas_data=atlas_data_dct,
im_vert=im_vert_data,
p_lst=p_lst)
if self.fname_ref is not None:
# Compute mean and std value in each labeled lesion
self._measure_within_im(im_lesion=im_lesion_data,
im_ref=Image(self.fname_ref).data,
label_lst=label_lst)
def register_data(im_src,
im_dest,
param_reg,
path_copy_warp=None,
rm_tmp=True):
'''
Parameters
----------
im_src: class Image: source image
im_dest: class Image: destination image
param_reg: str: registration parameter
path_copy_warp: path: path to copy the warping fields
Returns: im_src_reg: class Image: source image registered on destination image
-------
'''
# im_src and im_dest are already preprocessed (in theory: im_dest = mean_image)
# binarize images to get seg
im_src_seg = binarize(im_src, thr_min=1, thr_max=1)
im_dest_seg = binarize(im_dest)
# create tmp dir and go in it
tmp_dir = tmp_create()
curdir = os.getcwd()
os.chdir(tmp_dir)
# save image and seg
fname_src = 'src.nii.gz'
im_src.save(fname_src)
fname_src_seg = 'src_seg.nii.gz'
im_src_seg.save(fname_src_seg)
fname_dest = 'dest.nii.gz'
im_dest.save(fname_dest)
fname_dest_seg = 'dest_seg.nii.gz'
im_dest_seg.save(fname_dest_seg)
# do registration using param_reg
sct_register_multimodal.main(args=[
'-i', fname_src, '-d', fname_dest, '-iseg', fname_src_seg, '-dseg',
fname_dest_seg, '-param', param_reg
])
# get registration result
fname_src_reg = add_suffix(fname_src, '_reg')
im_src_reg = Image(fname_src_reg)
# get out of tmp dir
os.chdir(curdir)
# copy warping fields
if path_copy_warp is not None and os.path.isdir(
os.path.abspath(path_copy_warp)):
path_copy_warp = os.path.abspath(path_copy_warp)
file_src = extract_fname(fname_src)[1]
file_dest = extract_fname(fname_dest)[1]
fname_src2dest = 'warp_' + file_src + '2' + file_dest + '.nii.gz'
fname_dest2src = 'warp_' + file_dest + '2' + file_src + '.nii.gz'
copy(os.path.join(tmp_dir, fname_src2dest), path_copy_warp)
copy(os.path.join(tmp_dir, fname_dest2src), path_copy_warp)
if rm_tmp:
# remove tmp dir
rmtree(tmp_dir)
# return res image
return im_src_reg, fname_src2dest, fname_dest2src
def main(args=None):
"""
Main function
:param args:
:return:
"""
dim_list = ['x', 'y', 'z', 't']
# Get parser args
if args is None:
args = None if sys.argv[1:] else ['--help']
parser = get_parser()
arguments = parser.parse_args(args=args)
fname_in = arguments.i
fname_out = arguments.o
verbose = arguments.v
sct.init_sct(log_level=verbose, update=True) # Update log level
if '-type' in arguments:
output_type = arguments.type
else:
output_type = None
# Open file(s)
im = Image(fname_in)
data = im.data # 3d or 4d numpy array
dim = im.dim
# run command
if arguments.otsu is not None:
param = arguments.otsu
data_out = otsu(data, param)
elif arguments.adap is not None:
param = convert_list_str(arguments.adap, "int")
data_out = adap(data, param[0], param[1])
elif arguments.otsu_median is not None:
param = convert_list_str(arguments.otsu_median, "int")
data_out = otsu_median(data, param[0], param[1])
elif arguments.thr is not None:
param = arguments.thr
data_out = threshold(data, param)
elif arguments.percent is not None:
param = arguments.percent
data_out = perc(data, param)
elif arguments.bin is not None:
bin_thr = arguments.bin
data_out = binarise(data, bin_thr=bin_thr)
elif arguments.add is not None:
from numpy import sum
data2 = get_data_or_scalar(arguments.add, data)
data_concat = concatenate_along_4th_dimension(data, data2)
data_out = sum(data_concat, axis=3)
elif arguments.sub is not None:
data2 = get_data_or_scalar(arguments.sub, data)
data_out = data - data2
elif arguments.laplacian is not None:
sigmas = convert_list_str(arguments.laplacian, "float")
if len(sigmas) == 1:
sigmas = [sigmas for i in range(len(data.shape))]
elif len(sigmas) != len(data.shape):
printv(parser.usage.generate(error='ERROR: -laplacian need the same number of inputs as the number of image dimension OR only one input'))
# adjust sigma based on voxel size
sigmas = [sigmas[i] / dim[i + 4] for i in range(3)]
# smooth data
data_out = laplacian(data, sigmas)
elif arguments.mul is not None:
from numpy import prod
data2 = get_data_or_scalar(arguments.mul, data)
data_concat = concatenate_along_4th_dimension(data, data2)
data_out = prod(data_concat, axis=3)
elif arguments.div is not None:
from numpy import divide
data2 = get_data_or_scalar(arguments.div, data)
data_out = divide(data, data2)
elif arguments.mean is not None:
from numpy import mean
dim = dim_list.index(arguments.mean)
if dim + 1 > len(np.shape(data)): # in case input volume is 3d and dim=t
data = data[..., np.newaxis]
data_out = mean(data, dim)
elif arguments.rms is not None:
from numpy import mean, sqrt, square
dim = dim_list.index(arguments.rms)
if dim + 1 > len(np.shape(data)): # in case input volume is 3d and dim=t
data = data[..., np.newaxis]
data_out = sqrt(mean(square(data.astype(float)), dim))
elif arguments.std is not None:
from numpy import std
dim = dim_list.index(arguments.std)
if dim + 1 > len(np.shape(data)): # in case input volume is 3d and dim=t
data = data[..., np.newaxis]
data_out = std(data, dim, ddof=1)
elif arguments.smooth is not None:
sigmas = convert_list_str(arguments.smooth, "float")
if len(sigmas) == 1:
sigmas = [sigmas[0] for i in range(len(data.shape))]
elif len(sigmas) != len(data.shape):
printv(parser.usage.generate(error='ERROR: -smooth need the same number of inputs as the number of image dimension OR only one input'))
# adjust sigma based on voxel size
sigmas = [sigmas[i] / dim[i + 4] for i in range(3)]
# smooth data
data_out = smooth(data, sigmas)
elif arguments.dilate is not None:
data_out = dilate(data, convert_list_str(arguments.dilate, "int"))
elif arguments.erode is not None:
data_out = erode(data, convert_list_str(arguments.erode))
elif arguments.denoise is not None:
# parse denoising arguments
p, b = 1, 5 # default arguments
list_denoise = (arguments.denoise).split(",")
for i in list_denoise:
if 'p' in i:
p = int(i.split('=')[1])
if 'b' in i:
b = int(i.split('=')[1])
data_out = denoise_nlmeans(data, patch_radius=p, block_radius=b)
elif arguments.symmetrize is not None:
data_out = (data + data[list(range(data.shape[0] - 1, -1, -1)), :, :]) / float(2)
elif arguments.mi is not None:
# input 1 = from flag -i --> im
# input 2 = from flag -mi
im_2 = Image(arguments.mi)
compute_similarity(im.data, im_2.data, fname_out, metric='mi', verbose=verbose)
data_out = None
elif arguments.minorm is not None:
im_2 = Image(arguments.minorm)
compute_similarity(im.data, im_2.data, fname_out, metric='minorm', verbose=verbose)
data_out = None
elif arguments.corr is not None:
# input 1 = from flag -i --> im
# input 2 = from flag -mi
im_2 = Image(arguments.corr)
compute_similarity(im.data, im_2.data, fname_out, metric='corr', verbose=verbose)
data_out = None
# if no flag is set
else:
data_out = None
printv(parser.usage.generate(error='ERROR: you need to specify an operation to do on the input image'))
if data_out is not None:
# Write output
nii_out = Image(fname_in) # use header of input file
nii_out.data = data_out
nii_out.save(fname_out, dtype=output_type)
# TODO: case of multiple outputs
# assert len(data_out) == n_out
# if n_in == n_out:
# for im_in, d_out, fn_out in zip(nii, data_out, fname_out):
# im_in.data = d_out
# im_in.absolutepath = fn_out
# if "-w" in arguments:
# im_in.hdr.set_intent('vector', (), '')
# im_in.save()
# elif n_out == 1:
# nii[0].data = data_out[0]
# nii[0].absolutepath = fname_out[0]
# if "-w" in arguments:
# nii[0].hdr.set_intent('vector', (), '')
# nii[0].save()
# elif n_out > n_in:
# for dat_out, name_out in zip(data_out, fname_out):
# im_out = nii[0].copy()
# im_out.data = dat_out
# im_out.absolutepath = name_out
# if "-w" in arguments:
# im_out.hdr.set_intent('vector', (), '')
# im_out.save()
# else:
# printv(parser.usage.generate(error='ERROR: not the correct numbers of inputs and outputs'))
# display message
if data_out is not None:
sct.display_viewer_syntax([fname_out], verbose=verbose)
else:
printv('\nDone! File created: ' + fname_out, verbose, 'info')
def main(args=None):
parser = get_parser()
if args:
arguments = parser.parse_args(args)
else:
arguments = parser.parse_args(args=None if sys.argv[1:] else ['--help'])
verbosity = arguments.v
init_sct(log_level=verbosity, update=True) # Update log level
input_filename = arguments.i
output_fname = arguments.o
img = Image(input_filename)
dtype = None
if arguments.add is not None:
value = arguments.add
out = sct_labels.add(img, value)
elif arguments.create is not None:
labels = arguments.create
out = sct_labels.create_labels_empty(img, labels)
elif arguments.create_add is not None:
labels = arguments.create_add
out = sct_labels.create_labels(img, labels)
elif arguments.create_seg is not None:
labels = arguments.create_seg
out = sct_labels.create_labels_along_segmentation(img, labels)
elif arguments.cubic_to_point:
out = sct_labels.cubic_to_point(img)
elif arguments.display:
display_voxel(img, verbosity)
return
elif arguments.increment:
out = sct_labels.increment_z_inverse(img)
elif arguments.disc is not None:
ref = Image(arguments.disc)
out = sct_labels.labelize_from_discs(img, ref)
elif arguments.vert_body is not None:
levels = arguments.vert_body
if len(levels) == 1 and levels[0] == 0:
levels = None # all levels
out = sct_labels.label_vertebrae(img, levels)
elif arguments.vert_continuous:
out = sct_labels.continuous_vertebral_levels(img)
dtype = 'float32'
elif arguments.MSE is not None:
ref = Image(arguments.MSE)
mse = sct_labels.compute_mean_squared_error(img, ref)
printv(f"Computed MSE: {mse}")
return
elif arguments.remove_reference is not None:
ref = Image(arguments.remove_reference)
out = sct_labels.remove_missing_labels(img, ref)
elif arguments.remove_sym is not None:
# first pass use img as source
ref = Image(arguments.remove_reference)
out = sct_labels.remove_missing_labels(img, ref)
# second pass use previous pass result as reference
ref_out = sct_labels.remove_missing_labels(ref, out)
ref_out.save(path=ref.absolutepath)
elif arguments.remove is not None:
labels = arguments.remove
out = sct_labels.remove_labels_from_image(img, labels)
elif arguments.keep is not None:
labels = arguments.keep
out = sct_labels.remove_other_labels_from_image(img, labels)
elif arguments.create_viewer is not None:
msg = "" if arguments.msg is None else f"{arguments.msg}\n"
if arguments.ilabel is not None:
input_labels_img = Image(arguments.ilabel)
out = launch_manual_label_gui(img, input_labels_img, parse_num_list(arguments.create_viewer), msg)
else:
out = launch_sagittal_viewer(img, parse_num_list(arguments.create_viewer), msg)
printv("Generating output files...")
out.save(path=output_fname, dtype=dtype)
display_viewer_syntax([input_filename, output_fname])
if arguments.qc is not None:
generate_qc(fname_in1=input_filename, fname_seg=output_fname, args=args,
path_qc=os.path.abspath(arguments.qc), dataset=arguments.qc_dataset,
subject=arguments.qc_subject, process='sct_label_utils')
def pre_processing(fname_target, fname_sc_seg, fname_level=None, fname_manual_gmseg=None, new_res=0.3, square_size_size_mm=22.5, denoising=True, verbose=1, rm_tmp=True, for_model=False):
printv('\nPre-process data...', verbose, 'normal')
tmp_dir = tmp_create()
copy(fname_target, tmp_dir)
fname_target = ''.join(extract_fname(fname_target)[1:])
copy(fname_sc_seg, tmp_dir)
fname_sc_seg = ''.join(extract_fname(fname_sc_seg)[1:])
curdir = os.getcwd()
os.chdir(tmp_dir)
original_info = {'orientation': None, 'im_sc_seg_rpi': None, 'interpolated_images': []}
im_target = Image(fname_target).copy()
im_sc_seg = Image(fname_sc_seg).copy()
# get original orientation
printv(' Reorient...', verbose, 'normal')
original_info['orientation'] = im_target.orientation
# assert images are in the same orientation
assert im_target.orientation == im_sc_seg.orientation, "ERROR: the image to segment and it's SC segmentation are not in the same orientation"
im_target_rpi = im_target.copy().change_orientation('RPI', generate_path=True).save()
im_sc_seg_rpi = im_sc_seg.copy().change_orientation('RPI', generate_path=True).save()
original_info['im_sc_seg_rpi'] = im_sc_seg_rpi.copy() # target image in RPI will be used to post-process segmentations
# denoise using P. Coupe non local means algorithm (see [Manjon et al. JMRI 2010]) implemented in dipy
if denoising:
printv(' Denoise...', verbose, 'normal')
# crop image before denoising to fasten denoising
nx, ny, nz, nt, px, py, pz, pt = im_target_rpi.dim
size_x, size_y = (square_size_size_mm + 1) / px, (square_size_size_mm + 1) / py
size = int(np.ceil(max(size_x, size_y)))
# create mask
fname_mask = 'mask_pre_crop.nii.gz'
sct_create_mask.main(['-i', im_target_rpi.absolutepath, '-p', 'centerline,' + im_sc_seg_rpi.absolutepath, '-f', 'box', '-size', str(size), '-o', fname_mask])
# crop image
cropper = ImageCropper(im_target_rpi)
cropper.get_bbox_from_mask(Image(fname_mask))
im_target_rpi_crop = cropper.crop()
# crop segmentation
cropper = ImageCropper(im_sc_seg_rpi)
cropper.get_bbox_from_mask(Image(fname_mask))
im_sc_seg_rpi_crop = cropper.crop()
# denoising
from spinalcordtoolbox.math import denoise_nlmeans
block_radius = 3
block_radius = int(im_target_rpi_crop.data.shape[2] / 2) if im_target_rpi_crop.data.shape[2] < (block_radius * 2) else block_radius
patch_radius = block_radius - 1
data_denoised = denoise_nlmeans(im_target_rpi_crop.data, block_radius=block_radius, patch_radius=patch_radius)
im_target_rpi_crop.data = data_denoised
im_target_rpi = im_target_rpi_crop
im_sc_seg_rpi = im_sc_seg_rpi_crop
else:
fname_mask = None
# interpolate image to reference square image (resample and square crop centered on SC)
printv(' Interpolate data to the model space...', verbose, 'normal')
list_im_slices = interpolate_im_to_ref(im_target_rpi, im_sc_seg_rpi, new_res=new_res, sq_size_size_mm=square_size_size_mm)
original_info['interpolated_images'] = list_im_slices # list of images (not Slice() objects)
printv(' Mask data using the spinal cord segmentation...', verbose, 'normal')
list_sc_seg_slices = interpolate_im_to_ref(im_sc_seg_rpi, im_sc_seg_rpi, new_res=new_res, sq_size_size_mm=square_size_size_mm, interpolation_mode=1)
for i in range(len(list_im_slices)):
# list_im_slices[i].data[list_sc_seg_slices[i].data == 0] = 0
list_sc_seg_slices[i] = binarize(list_sc_seg_slices[i], thr_min=0.5, thr_max=1)
list_im_slices[i].data = list_im_slices[i].data * list_sc_seg_slices[i].data
printv(' Split along rostro-caudal direction...', verbose, 'normal')
list_slices_target = [Slice(slice_id=i, im=im_slice.data, gm_seg=[], wm_seg=[]) for i, im_slice in enumerate(list_im_slices)]
# load vertebral levels
if fname_level is not None:
printv(' Load vertebral levels...', verbose, 'normal')
# copy level file to tmp dir
os.chdir(curdir)
copy(fname_level, tmp_dir)
os.chdir(tmp_dir)
# change fname level to only file name (path = tmp dir now)
fname_level = ''.join(extract_fname(fname_level)[1:])
# load levels
list_slices_target = load_level(list_slices_target, fname_level)
os.chdir(curdir)
# load manual gmseg if there is one (model data)
if fname_manual_gmseg is not None:
printv('\n\tLoad manual GM segmentation(s) ...', verbose, 'normal')
list_slices_target = load_manual_gmseg(list_slices_target, fname_manual_gmseg, tmp_dir, im_sc_seg_rpi, new_res, square_size_size_mm, for_model=for_model, fname_mask=fname_mask)
if rm_tmp:
# remove tmp folder
rmtree(tmp_dir)
return list_slices_target, original_info
def main(argv=None):
parser = get_parser()
arguments = parser.parse_args(argv)
verbose = arguments.v
set_loglevel(verbose=verbose)
fname_input1 = arguments.i
fname_input2 = arguments.d
tmp_dir = tmp_create() # create tmp directory
tmp_dir = os.path.abspath(tmp_dir)
# copy input files to tmp directory
fname_input1_tmp = 'tmp1_' + ''.join(extract_fname(fname_input1)[1:])
fname_input2_tmp = 'tmp2_' + ''.join(extract_fname(fname_input2)[1:])
copy(fname_input1, os.path.join(tmp_dir, fname_input1_tmp))
copy(fname_input2, os.path.join(tmp_dir, fname_input2_tmp))
fname_input1 = fname_input1_tmp
fname_input2 = fname_input2_tmp
curdir = os.getcwd()
os.chdir(tmp_dir) # go to tmp directory
im_1 = Image(fname_input1)
im_2 = Image(fname_input2)
if arguments.bin is not None:
im_1.data = binarize(im_1.data, 0)
fname_input1_bin = add_suffix(fname_input1, '_bin')
im_1.save(fname_input1_bin, mutable=True)
im_2.data = binarize(im_2.data, 0)
fname_input2_bin = add_suffix(fname_input2, '_bin')
im_2.save(fname_input2_bin, mutable=True)
# Use binarized images in subsequent steps
fname_input1 = fname_input1_bin
fname_input2 = fname_input2_bin
# copy header of im_1 to im_2
im_2.header = im_1.header
im_2.save()
cmd = ['isct_dice_coefficient', fname_input1, fname_input2]
if vars(arguments)["2d_slices"] is not None:
cmd += ['-2d-slices', str(vars(arguments)["2d_slices"])]
if arguments.b is not None:
bounding_box = arguments.b
cmd += ['-b'] + bounding_box
if arguments.bmax is not None and arguments.bmax == 1:
cmd += ['-bmax']
if arguments.bzmax is not None and arguments.bzmax == 1:
cmd += ['-bzmax']
if arguments.o is not None:
path_output, fname_output, ext = extract_fname(arguments.o)
cmd += ['-o', fname_output + ext]
rm_tmp = bool(arguments.r)
# # Computation of Dice coefficient using Python implementation.
# # commented for now as it does not cover all the feature of isct_dice_coefficient
# #from spinalcordtoolbox.image import Image, compute_dice
# #dice = compute_dice(Image(fname_input1), Image(fname_input2), mode='3d', zboundaries=False)
# #printv('Dice (python-based) = ' + str(dice), verbose)
status, output = run_proc(cmd, verbose, is_sct_binary=True)
os.chdir(curdir) # go back to original directory
# copy output file into original directory
if arguments.o is not None:
copy(os.path.join(tmp_dir, fname_output + ext),
os.path.join(path_output, fname_output + ext))
# remove tmp_dir
if rm_tmp:
rmtree(tmp_dir)
printv(output, verbose)
def load_level(list_slices_target, fname_level):
verbose = 1
path_level, file_level, ext_level = extract_fname(fname_level)
# ####### Check if the level file is an image or a text file
# Level file is an image
if ext_level in ['.nii', '.nii.gz']:
im_level = Image(fname_level).change_orientation("IRP")
list_level = []
list_med_level = []
for slice_level in im_level.data:
try:
# vertebral level of the slice
l = np.mean(slice_level[slice_level > 0])
# median of the vertebral level of the slice: if all voxels are int, med will be an int.
med = np.median(slice_level[slice_level > 0])
# change med in int if it is an int
med = int(med) if int(med) == med else med
except Exception as e:
printv('WARNING: ' + str(e) + '\nNo level label found. Level will be set to 0 for this slice', verbose, 'warning')
l = 0
med = 0
list_level.append(l)
list_med_level.append(med)
# if all median of level are int for all slices : consider level as int
if all([isinstance(med, int) for med in list_med_level]):
# level as int are placed in the middle of each vertebra (that's why there is a "+0.5")
list_level = [int(np.round(l)) + 0.5 for l in list_level]
# Level file is a text file
elif ext_level == '.txt':
file_level = open(fname_level, 'r')
lines_level = file_level.readlines()
file_level.close()
list_level_by_slice = []
list_type_level = [] # True or int, False for float
for line in lines_level[1:]:
i_slice, level = line.split(',')
# correct level value
for c in [' ', '\n', '\r', '\t']:
level = level.replace(c, '')
try:
level = float(level)
except Exception as e:
# adapt if level value is not unique
if len(level) > 2:
l1 = l2 = 0
if '-' in level:
l1, l2 = level.split('-')
elif '/' in level:
l1, l2 = level.split('/')
# convention = the vertebral disk between two levels belong to the lower level (C2-C3 = C3)
level = max(float(l1), float(l2))
else:
# level unrecognized
level = 0
i_slice = int(i_slice)
list_type_level.append(int(level) == level)
list_level_by_slice.append((i_slice, level))
# sort list by number of slice
list_level_by_slice.sort()
if all(list_type_level): # levels are int
# add 0.5 to the int level to place in the middle of the vertebra
to_add = 0.5
else:
# levels are float: keep them untouched
to_add = 0
list_level = [l[1] + to_add for l in list_level_by_slice]
# Level file is not recognized
else:
list_level = None
printv('ERROR: the level file is nor an image nor a text file ...', verbose, 'error')
# ####### Set level number for each slice of list_slices_target:
for target_slice, level in zip(list_slices_target, list_level):
target_slice.set(level=level)
return list_slices_target
def main(argv=None):
"""
Main function
:param argv:
:return:
"""
parser = get_parser()
arguments = parser.parse_args(argv)
verbose = arguments.v
set_global_loglevel(verbose=verbose)
dim_list = ['x', 'y', 'z', 't']
fname_in = arguments.i
fname_out = arguments.o
output_type = arguments.type
# Open file(s)
im = Image(fname_in)
data = im.data # 3d or 4d numpy array
dim = im.dim
# run command
if arguments.otsu is not None:
param = arguments.otsu
data_out = sct_math.otsu(data, param)
elif arguments.adap is not None:
param = arguments.adap
data_out = sct_math.adap(data, param[0], param[1])
elif arguments.otsu_median is not None:
param = arguments.otsu_median
data_out = sct_math.otsu_median(data, param[0], param[1])
elif arguments.thr is not None:
param = arguments.thr
data_out = sct_math.threshold(data, param)
elif arguments.percent is not None:
param = arguments.percent
data_out = sct_math.perc(data, param)
elif arguments.bin is not None:
bin_thr = arguments.bin
data_out = sct_math.binarize(data, bin_thr=bin_thr)
elif arguments.add is not None:
data2 = get_data_or_scalar(arguments.add, data)
data_concat = sct_math.concatenate_along_4th_dimension(data, data2)
data_out = np.sum(data_concat, axis=3)
elif arguments.sub is not None:
data2 = get_data_or_scalar(arguments.sub, data)
data_out = data - data2
elif arguments.laplacian is not None:
sigmas = arguments.laplacian
if len(sigmas) == 1:
sigmas = [sigmas for i in range(len(data.shape))]
elif len(sigmas) != len(data.shape):
printv(
parser.error(
'ERROR: -laplacian need the same number of inputs as the number of image dimension OR only one input'
))
# adjust sigma based on voxel size
sigmas = [sigmas[i] / dim[i + 4] for i in range(3)]
# smooth data
data_out = sct_math.laplacian(data, sigmas)
elif arguments.mul is not None:
data2 = get_data_or_scalar(arguments.mul, data)
data_concat = sct_math.concatenate_along_4th_dimension(data, data2)
data_out = np.prod(data_concat, axis=3)
elif arguments.div is not None:
data2 = get_data_or_scalar(arguments.div, data)
data_out = np.divide(data, data2)
elif arguments.mean is not None:
dim = dim_list.index(arguments.mean)
if dim + 1 > len(
np.shape(data)): # in case input volume is 3d and dim=t
data = data[..., np.newaxis]
data_out = np.mean(data, dim)
elif arguments.rms is not None:
dim = dim_list.index(arguments.rms)
if dim + 1 > len(
np.shape(data)): # in case input volume is 3d and dim=t
data = data[..., np.newaxis]
data_out = np.sqrt(np.mean(np.square(data.astype(float)), dim))
elif arguments.std is not None:
dim = dim_list.index(arguments.std)
if dim + 1 > len(
np.shape(data)): # in case input volume is 3d and dim=t
data = data[..., np.newaxis]
data_out = np.std(data, dim, ddof=1)
elif arguments.smooth is not None:
sigmas = arguments.smooth
if len(sigmas) == 1:
sigmas = [sigmas[0] for i in range(len(data.shape))]
elif len(sigmas) != len(data.shape):
printv(
parser.error(
'ERROR: -smooth need the same number of inputs as the number of image dimension OR only one input'
))
# adjust sigma based on voxel size
sigmas = [sigmas[i] / dim[i + 4] for i in range(3)]
# smooth data
data_out = sct_math.smooth(data, sigmas)
elif arguments.dilate is not None:
if arguments.shape in ['disk', 'square'] and arguments.dim is None:
printv(
parser.error(
'ERROR: -dim is required for -dilate with 2D morphological kernel'
))
data_out = sct_math.dilate(data,
size=arguments.dilate,
shape=arguments.shape,
dim=arguments.dim)
elif arguments.erode is not None:
if arguments.shape in ['disk', 'square'] and arguments.dim is None:
printv(
parser.error(
'ERROR: -dim is required for -erode with 2D morphological kernel'
))
data_out = sct_math.erode(data,
size=arguments.erode,
shape=arguments.shape,
dim=arguments.dim)
elif arguments.denoise is not None:
# parse denoising arguments
p, b = 1, 5 # default arguments
list_denoise = (arguments.denoise).split(",")
for i in list_denoise:
if 'p' in i:
p = int(i.split('=')[1])
if 'b' in i:
b = int(i.split('=')[1])
data_out = sct_math.denoise_nlmeans(data,
patch_radius=p,
block_radius=b)
elif arguments.symmetrize is not None:
data_out = (data + data[list(range(data.shape[0] -
1, -1, -1)), :, :]) / float(2)
elif arguments.mi is not None:
# input 1 = from flag -i --> im
# input 2 = from flag -mi
im_2 = Image(arguments.mi)
compute_similarity(im,
im_2,
fname_out,
metric='mi',
metric_full='Mutual information',
verbose=verbose)
data_out = None
elif arguments.minorm is not None:
im_2 = Image(arguments.minorm)
compute_similarity(im,
im_2,
fname_out,
metric='minorm',
metric_full='Normalized Mutual information',
verbose=verbose)
data_out = None
elif arguments.corr is not None:
# input 1 = from flag -i --> im
# input 2 = from flag -mi
im_2 = Image(arguments.corr)
compute_similarity(im,
im_2,
fname_out,
metric='corr',
metric_full='Pearson correlation coefficient',
verbose=verbose)
data_out = None
# if no flag is set
else:
data_out = None
printv(
parser.error(
'ERROR: you need to specify an operation to do on the input image'
))
if data_out is not None:
# Write output
nii_out = Image(fname_in) # use header of input file
nii_out.data = data_out
nii_out.save(fname_out, dtype=output_type)
# TODO: case of multiple outputs
# assert len(data_out) == n_out
# if n_in == n_out:
# for im_in, d_out, fn_out in zip(nii, data_out, fname_out):
# im_in.data = d_out
# im_in.absolutepath = fn_out
# if arguments.w is not None:
# im_in.hdr.set_intent('vector', (), '')
# im_in.save()
# elif n_out == 1:
# nii[0].data = data_out[0]
# nii[0].absolutepath = fname_out[0]
# if arguments.w is not None:
# nii[0].hdr.set_intent('vector', (), '')
# nii[0].save()
# elif n_out > n_in:
# for dat_out, name_out in zip(data_out, fname_out):
# im_out = nii[0].copy()
# im_out.data = dat_out
# im_out.absolutepath = name_out
# if arguments.w is not None:
# im_out.hdr.set_intent('vector', (), '')
# im_out.save()
# else:
# printv(parser.usage.generate(error='ERROR: not the correct numbers of inputs and outputs'))
# display message
if data_out is not None:
display_viewer_syntax([fname_out], verbose=verbose)
else:
printv('\nDone! File created: ' + fname_out, verbose, 'info')
def compute_dti(fname_in, fname_bvals, fname_bvecs, prefix, method, evecs,
file_mask, verbose):
"""
Compute DTI.
:param fname_in: input 4d file.
:param bvals: bvals txt file
:param bvecs: bvecs txt file
:param prefix: output prefix. Example: "dti_"
:param method: algo for computing dti
:param evecs: bool: output diffusion tensor eigenvectors and eigenvalues
:return: True/False
"""
# Open file.
from spinalcordtoolbox.image import Image
nii = Image(fname_in)
data = nii.data
printv('data.shape (%d, %d, %d, %d)' % data.shape)
# open bvecs/bvals
from dipy.io import read_bvals_bvecs
bvals, bvecs = read_bvals_bvecs(fname_bvals, fname_bvecs)
from dipy.core.gradients import gradient_table
gtab = gradient_table(bvals, bvecs)
# mask and crop the data. This is a quick way to avoid calculating Tensors on the background of the image.
if not file_mask == '':
printv('Open mask file...', verbose)
# open mask file
nii_mask = Image(file_mask)
mask = nii_mask.data
# fit tensor model
printv('Computing tensor using "' + method + '" method...', verbose)
import dipy.reconst.dti as dti
if method == 'standard':
tenmodel = dti.TensorModel(gtab)
if file_mask == '':
tenfit = tenmodel.fit(data)
else:
tenfit = tenmodel.fit(data, mask)
elif method == 'restore':
import dipy.denoise.noise_estimate as ne
sigma = ne.estimate_sigma(data)
dti_restore = dti.TensorModel(gtab, fit_method='RESTORE', sigma=sigma)
if file_mask == '':
tenfit = dti_restore.fit(data)
else:
tenfit = dti_restore.fit(data, mask)
# Compute metrics
printv('Computing metrics...', verbose)
# FA
nii.data = tenfit.fa
nii.save(prefix + 'FA.nii.gz', dtype='float32')
# MD
nii.data = tenfit.md
nii.save(prefix + 'MD.nii.gz', dtype='float32')
# RD
nii.data = tenfit.rd
nii.save(prefix + 'RD.nii.gz', dtype='float32')
# AD
nii.data = tenfit.ad
nii.save(prefix + 'AD.nii.gz', dtype='float32')
if evecs:
data_evecs = tenfit.evecs
data_evals = tenfit.evals
# output 1st (V1), 2nd (V2) and 3rd (V3) eigenvectors as 4d data
for idim in range(3):
nii.data = data_evecs[:, :, :, :, idim]
nii.save(prefix + 'V' + str(idim + 1) + '.nii.gz', dtype="float32")
nii.data = data_evals[:, :, :, idim]
nii.save(prefix + 'E' + str(idim + 1) + '.nii.gz', dtype="float32")
return True
def propseg(img_input, options_dict):
"""
:param img_input: source image, to be segmented
:param options_dict: arguments as dictionary
:return: segmented Image
"""
arguments = options_dict
fname_input_data = img_input.absolutepath
fname_data = os.path.abspath(fname_input_data)
contrast_type = arguments.c
contrast_type_conversion = {
't1': 't1',
't2': 't2',
't2s': 't2',
'dwi': 't1'
}
contrast_type_propseg = contrast_type_conversion[contrast_type]
# Starting building the command
cmd = ['isct_propseg', '-t', contrast_type_propseg]
if arguments.o is not None:
fname_out = arguments.o
else:
fname_out = os.path.basename(add_suffix(fname_data, "_seg"))
folder_output = str(pathlib.Path(fname_out).parent)
cmd += ['-o', folder_output]
if not os.path.isdir(folder_output) and os.path.exists(folder_output):
logger.error("output directory %s is not a valid directory" %
folder_output)
if not os.path.exists(folder_output):
os.makedirs(folder_output)
if arguments.down is not None:
cmd += ["-down", str(arguments.down)]
if arguments.up is not None:
cmd += ["-up", str(arguments.up)]
remove_temp_files = arguments.r
verbose = int(arguments.v)
# Update for propseg binary
if verbose > 0:
cmd += ["-verbose"]
# Output options
if arguments.mesh is not None:
cmd += ["-mesh"]
if arguments.centerline_binary is not None:
cmd += ["-centerline-binary"]
if arguments.CSF is not None:
cmd += ["-CSF"]
if arguments.centerline_coord is not None:
cmd += ["-centerline-coord"]
if arguments.cross is not None:
cmd += ["-cross"]
if arguments.init_tube is not None:
cmd += ["-init-tube"]
if arguments.low_resolution_mesh is not None:
cmd += ["-low-resolution-mesh"]
# TODO: Not present. Why is this here? Was this renamed?
# if arguments.detect_nii is not None:
# cmd += ["-detect-nii"]
# TODO: Not present. Why is this here? Was this renamed?
# if arguments.detect_png is not None:
# cmd += ["-detect-png"]
# Helping options
use_viewer = None
use_optic = True # enabled by default
init_option = None
rescale_header = arguments.rescale
if arguments.init is not None:
init_option = float(arguments.init)
if init_option < 0:
printv(
'Command-line usage error: ' + str(init_option) +
" is not a valid value for '-init'", 1, 'error')
sys.exit(1)
if arguments.init_centerline is not None:
if str(arguments.init_centerline) == "viewer":
use_viewer = "centerline"
elif str(arguments.init_centerline) == "hough":
use_optic = False
else:
if rescale_header is not 1:
fname_labels_viewer = func_rescale_header(str(
arguments.init_centerline),
rescale_header,
verbose=verbose)
else:
fname_labels_viewer = str(arguments.init_centerline)
cmd += ["-init-centerline", fname_labels_viewer]
use_optic = False
if arguments.init_mask is not None:
if str(arguments.init_mask) == "viewer":
use_viewer = "mask"
else:
if rescale_header is not 1:
fname_labels_viewer = func_rescale_header(
str(arguments.init_mask), rescale_header)
else:
fname_labels_viewer = str(arguments.init_mask)
cmd += ["-init-mask", fname_labels_viewer]
use_optic = False
if arguments.mask_correction is not None:
cmd += ["-mask-correction", str(arguments.mask_correction)]
if arguments.radius is not None:
cmd += ["-radius", str(arguments.radius)]
# TODO: Not present. Why is this here? Was this renamed?
# if arguments.detect_n is not None:
# cmd += ["-detect-n", str(arguments.detect_n)]
# TODO: Not present. Why is this here? Was this renamed?
# if arguments.detect_gap is not None:
# cmd += ["-detect-gap", str(arguments.detect_gap)]
# TODO: Not present. Why is this here? Was this renamed?
# if arguments.init_validation is not None:
# cmd += ["-init-validation"]
if arguments.nbiter is not None:
cmd += ["-nbiter", str(arguments.nbiter)]
if arguments.max_area is not None:
cmd += ["-max-area", str(arguments.max_area)]
if arguments.max_deformation is not None:
cmd += ["-max-deformation", str(arguments.max_deformation)]
if arguments.min_contrast is not None:
cmd += ["-min-contrast", str(arguments.min_contrast)]
if arguments.d is not None:
cmd += ["-d", str(arguments["-d"])]
if arguments.distance_search is not None:
cmd += ["-dsearch", str(arguments.distance_search)]
if arguments.alpha is not None:
cmd += ["-alpha", str(arguments.alpha)]
# check if input image is in 3D. Otherwise itk image reader will cut the 4D image in 3D volumes and only take the first one.
image_input = Image(fname_data)
image_input_rpi = image_input.copy().change_orientation('RPI')
nx, ny, nz, nt, px, py, pz, pt = image_input_rpi.dim
if nt > 1:
printv(
'ERROR: your input image needs to be 3D in order to be segmented.',
1, 'error')
path_data, file_data, ext_data = extract_fname(fname_data)
path_tmp = tmp_create(basename="label_vertebrae")
# rescale header (see issue #1406)
if rescale_header is not 1:
fname_data_propseg = func_rescale_header(fname_data, rescale_header)
else:
fname_data_propseg = fname_data
# add to command
cmd += ['-i', fname_data_propseg]
# if centerline or mask is asked using viewer
if use_viewer:
from spinalcordtoolbox.gui.base import AnatomicalParams
from spinalcordtoolbox.gui.centerline import launch_centerline_dialog
params = AnatomicalParams()
if use_viewer == 'mask':
params.num_points = 3
params.interval_in_mm = 15 # superior-inferior interval between two consecutive labels
params.starting_slice = 'midfovminusinterval'
if use_viewer == 'centerline':
# setting maximum number of points to a reasonable value
params.num_points = 20
params.interval_in_mm = 30
params.starting_slice = 'top'
im_data = Image(fname_data_propseg)
im_mask_viewer = zeros_like(im_data)
# im_mask_viewer.absolutepath = add_suffix(fname_data_propseg, '_labels_viewer')
controller = launch_centerline_dialog(im_data, im_mask_viewer, params)
fname_labels_viewer = add_suffix(fname_data_propseg, '_labels_viewer')
if not controller.saved:
printv(
'The viewer has been closed before entering all manual points. Please try again.',
1, 'error')
sys.exit(1)
# save labels
controller.as_niftii(fname_labels_viewer)
# add mask filename to parameters string
if use_viewer == "centerline":
cmd += ["-init-centerline", fname_labels_viewer]
elif use_viewer == "mask":
cmd += ["-init-mask", fname_labels_viewer]
# If using OptiC
elif use_optic:
image_centerline = optic.detect_centerline(image_input, contrast_type,
verbose)
fname_centerline_optic = os.path.join(path_tmp,
'centerline_optic.nii.gz')
image_centerline.save(fname_centerline_optic)
cmd += ["-init-centerline", fname_centerline_optic]
if init_option is not None:
if init_option > 1:
init_option /= (nz - 1)
cmd += ['-init', str(init_option)]
# enabling centerline extraction by default (needed by check_and_correct_segmentation() )
cmd += ['-centerline-binary']
# run propseg
status, output = run_proc(cmd,
verbose,
raise_exception=False,
is_sct_binary=True)
# check status is not 0
if not status == 0:
printv(
'Automatic cord detection failed. Please initialize using -init-centerline or -init-mask (see help)',
1, 'error')
sys.exit(1)
# build output filename
fname_seg = os.path.join(folder_output, fname_out)
fname_centerline = os.path.join(
folder_output, os.path.basename(add_suffix(fname_data, "_centerline")))
# in case header was rescaled, we need to update the output file names by removing the "_rescaled"
if rescale_header is not 1:
mv(
os.path.join(
folder_output,
add_suffix(os.path.basename(fname_data_propseg), "_seg")),
fname_seg)
mv(
os.path.join(
folder_output,
add_suffix(os.path.basename(fname_data_propseg),
"_centerline")), fname_centerline)
# if user was used, copy the labelled points to the output folder (they will then be scaled back)
if use_viewer:
fname_labels_viewer_new = os.path.join(
folder_output,
os.path.basename(add_suffix(fname_data, "_labels_viewer")))
copy(fname_labels_viewer, fname_labels_viewer_new)
# update variable (used later)
fname_labels_viewer = fname_labels_viewer_new
# check consistency of segmentation
if arguments.correct_seg:
check_and_correct_segmentation(fname_seg,
fname_centerline,
folder_output=folder_output,
threshold_distance=3.0,
remove_temp_files=remove_temp_files,
verbose=verbose)
# copy header from input to segmentation to make sure qform is the same
printv("Copy header input --> output(s) to make sure qform is the same.",
verbose)
list_fname = [fname_seg, fname_centerline]
if use_viewer:
list_fname.append(fname_labels_viewer)
for fname in list_fname:
im = Image(fname)
im.header = image_input.header
im.save(dtype='int8'
) # they are all binary masks hence fine to save as int8
return Image(fname_seg)
