def main(args):
starttime = datetime.now()
parser = parsefn()
indir, outdir, maskimg, segment, chann, chanp, down, fwhm, noise, bins, levels, iters, thresh, mulpower, outnii, \
vx, vz, chan = parse_inputs(parser, args)
# make downsampled nii
createnii(indir, down, chann, chanp, chan, outnii, vx, vz)
# bias corr
totaliters = ('%dx' % iters) * levels
totaliters = totaliters[:-1]
hist = [fwhm, noise, bins]
conv = [totaliters, thresh]
niidir = 'niftis'
niiname = '%s/%s_%02dx_down_%s_chan.nii.gz' % (niidir, outnii, down, chan)
corname = os.path.basename(niiname).split('.nii')[0]
niicorr = os.path.join(niidir, '%s_corr.nii.gz' % corname)
field = os.path.join(niidir, '%s_biasfield.nii.gz' % corname)
mask = os.path.join(niidir, '%s_mask.nii.gz' % corname)
with add_paths():
biascorrnii(niiname, maskimg, segment, hist, conv, niicorr, mask, field)
# up-sample bias field & mask
biasnii = nib.load(field)
bias = biasnii.get_data()
biasres = scipy.ndimage.interpolation.zoom(bias, [1, 1, down])
if maskimg == 1:
masknii = nib.load(mask).get_data()
maskres = scipy.ndimage.interpolation.zoom(masknii, [1, 1, down], order=1)
else:
maskres = None
# sort files
if chanp is None:
file_list = sorted(glob.glob("%s/*.tif" % indir), key=numericalsort)
else:
file_list = sorted(glob.glob("%s/*%s%01d*.tif" % (indir, chanp, chann)), key=numericalsort)
cpuload = 0.95
cpus = multiprocessing.cpu_count()
ncpus = int(cpuload * cpus)
print("\n Correcting TIFF images in parallel using %02d cpus" % ncpus)
Parallel(n_jobs=ncpus, backend='threading')(
delayed(applycorr)(i, tif, outdir, biasres, down, mulpower, maskimg, maskres)
for i, tif in enumerate(file_list))
# print("\n Intensity correction done in %s ... Have a good day!\n" % (datetime.now() - starttime))
statement.main(['Intensity correction', '%s' % (datetime.now() - starttime)])
def main(args):
starttime = datetime.now()
parser = parsefn()
inlbls, img, metric, outfile, s = parse_inputs(parser, args)
# load img
print("Reading Input Labels")
nii = nib.load(img)
data = nii.get_data()
# read graph
print("Reading ARA ontology structure_graph")
miracl_home = os.environ['MIRACL_HOME']
arastrctcsv = "%s/atlases/ara/ara_mouse_structure_graph_hemi_combined.csv" % miracl_home
aragraph = pd.read_csv(arastrctcsv)
# get lbls
imglbls = getalllbls(data)
# flatten
dataflat = data.flatten()
lblvols = []
print("Computing volumes for input labels...")
for i, inlbl in enumerate(inlbls):
vol = computevolumes(aragraph, dataflat, inlbl, imglbls, metric)
lblvols.append(vol)
df = pd.DataFrame([inlbls, lblvols])
if s == 1:
df = df.T.sort_values(1, ascending=False)
else:
df = df.T
df.columns = ['Labels', 'Volumes (# of voxels)']
df.to_csv(outfile, index=False)
# print ("\n Parent labels volume computation done in %s ... Have a good day!\n" % (datetime.now() - starttime))
endstatement.main(task='Parent labels volume computation',
timediff='%s' % (datetime.now() - starttime))
def run_endstatmenet(parser, args):
miracl_utilfn_endstatement.main(args)