def _run_interface(self, runtime):
args = {}
for key in [k for k,_ in self.inputs.items() if k not in BaseInterfaceInputSpec().trait_names()]:
value = getattr(self.inputs, key)
if isdefined(value):
if key in ['mean_volume', 'reference_volume']:
nii = nb.load(value)
value = nii.get_data()
args[key] = value
brain_mask = compute_mask(**args)
self._brain_mask_path = os.path.abspath("brain_mask.nii")
nb.save(nb.Nifti1Image(brain_mask.astype(np.uint8), nii.get_affine()), self._brain_mask_path)
return runtime
def _run_interface(self, runtime):
from nipy.labs.mask import compute_mask
args = {}
for key in [k for k, _ in self.inputs.items() if k not in BaseInterfaceInputSpec().trait_names()]:
value = getattr(self.inputs, key)
if isdefined(value):
if key in ["mean_volume", "reference_volume"]:
nii = nb.load(value)
value = nii.get_data()
args[key] = value
brain_mask = compute_mask(**args)
_, name, ext = split_filename(self.inputs.mean_volume)
self._brain_mask_path = os.path.abspath("%s_mask.%s" % (name, ext))
nb.save(nb.Nifti1Image(brain_mask.astype(np.uint8), nii.get_affine()), self._brain_mask_path)
return runtime
def inter_run_mask(fnames, hist_m=.3, hist_M=.8):
"""
create the inter runs mask :
- first create a mask of all constant voxels on axis 3
- second create a mask of the mean
-
"""
EPSFCT = 10000. # factor: multiply numpy epsilon by it
imgs = [nib.load(fn) for fn in fnames]
# initialize mask to ones
mask = np.ones(imgs[0].shape[:3])
for img in imgs:
arrstd = np.asarray(img.get_data().std(axis=3))
mask = np.logical_and(mask, arrstd > EPSFCT*np.finfo(arrstd.dtype).eps)
mask_mean = compute_mask(img.get_data().mean(axis=3), None, \
m=hist_m, M=hist_M, cc=True, opening=2, exclude_zeros=False)
mask = np.logical_and(mask, mask_mean)
return mask
def _run_interface(self, runtime):
args = {}
for key in [k for k, _ in self.inputs.items()
if k not in BaseInterfaceInputSpec().trait_names()]:
value = getattr(self.inputs, key)
if isdefined(value):
if key in ['mean_volume', 'reference_volume']:
nii = nb.load(value)
value = nii.get_data()
args[key] = value
brain_mask = compute_mask(**args)
_, name, ext = split_filename(self.inputs.mean_volume)
self._brain_mask_path = os.path.abspath("%s_mask.%s" % (name, ext))
nb.save(nb.Nifti1Image(brain_mask.astype(np.uint8),
nii.get_affine()), self._brain_mask_path)
return runtime
def create_mask_pass1(img):
"""
This is the first pass in creating a mask. The skimage.exposure is used as it helps to get rid
of the glue around the tissue. This also calls the compute_mask method which is part
of the nipy package which you will need to install from github
:param img: the raw image
:return: the 1st pass of the image
"""
img = exposure.adjust_log(img, 1)
img = exposure.adjust_gamma(img, 2)
mask = compute_mask(img, m=0.2, M=0.9, cc=False, opening=2, exclude_zeros=True)
mask = mask.astype(int)
mask[mask==0] = 0
mask[mask==1] = 255
kernel = np.ones((5, 5), np.uint8)
mask = cv2.dilate(mask.astype(np.uint8), kernel, iterations=2)
mask = mask.astype(np.uint8)
return mask
def _run_interface(self, runtime):
from nipy.labs.mask import compute_mask
args = {}
for key in [
k for k, _ in list(self.inputs.items())
if k not in BaseInterfaceInputSpec().trait_names()
]:
value = getattr(self.inputs, key)
if isdefined(value):
if key in ["mean_volume", "reference_volume"]:
value = np.asanyarray(nb.load(value).dataobj)
args[key] = value
brain_mask = compute_mask(**args)
_, name, ext = split_filename(self.inputs.mean_volume)
self._brain_mask_path = os.path.abspath("%s_mask.%s" % (name, ext))
nb.save(
nb.Nifti1Image(brain_mask.astype(np.uint8), nii.affine),
self._brain_mask_path,
)
return runtime
cv2.imwrite(ch1_outpath, ch1_img.astype(np.uint8))
cv2.imwrite(ch2_outpath, ch2_img.astype(np.uint8))
cv2.imwrite(ch3_outpath, ch3_img.astype(np.uint8))
"""
max_width = 2000
max_height = 1000
for channel in [2]:
INPUT = os.path.join(BASEINPUT, f'CH{channel}/thumbnail')
files = sorted(os.listdir(INPUT))
for file in tqdm(files):
infile = os.path.join(INPUT, file)
img = io.imread(infile)
img, _ = remove_strip(img)
mask = compute_mask(img, m=0.2, M=0.9, cc=False, opening=2, exclude_zeros=True)
mask = mask.astype(int)
mask[mask==0] = 0
mask[mask==1] = 255
kernel = np.ones((5, 5), np.uint8)
mask = cv2.dilate(mask.astype(np.uint8), kernel, iterations=2)
mask = mask.astype(np.uint8)
fixed = cv2.bitwise_and(img, img, mask=mask)
fixed = equalized(fixed)
BASEOUTPUT = os.path.join(BASEINPUT, f'CH{channel}/thumbnail_cleaned')
outpath = os.path.join(BASEOUTPUT, file)
os.makedirs(BASEOUTPUT, exist_ok=True)
fixed = np.rot90(fixed, 3, axes=(1,0))
fixed = np.flip(fixed, axis=1)
fixed = place_image(fixed, file, max_width, max_height, 0)
