def segmentLesions3Dv2(pathInpNii,
dirWithModel,
pathOutNii=None,
outSize=None,
isDebug=False,
threshold=None,
path_lungs=None):
if isinstance(pathInpNii, str): # or isinstance(pathInpNii, unicode):
isInpFromFile = True
if not os.path.isfile(pathInpNii):
raise Exception('Cant find input file [%s]' % pathInpNii)
else:
isInpFromFile = False
if not os.path.isdir(dirWithModel):
raise Exception('Cant find directory with model [%s]' % dirWithModel)
if pathOutNii is not None:
outDir = os.path.dirname(os.path.abspath(pathOutNii))
if not os.path.isdir(outDir):
raise Exception(
'Cant find output directory [%s], create directory for output file before this call'
% outDir)
batcherInfer = InferencerLesion3Dv2()
batcherInfer.load_model(path_model=dirWithModel)
if isDebug:
batcherInfer.model.summary()
ret = batcherInfer.inference([pathInpNii], batchSize=1)
outMsk = ret[0]
if isInpFromFile:
tmpNii = nib.load(pathInpNii)
else:
tmpNii = pathInpNii
#
outMskNii = nib.Nifti1Image(outMsk.copy().astype(np.uint8),
tmpNii.affine,
header=tmpNii.header)
if outSize is not None:
outMskNii = resizeNii(outMskNii, newSize=outSize, parOrder=0)
if path_lungs is not None:
tmp_affine = outMskNii.affine
tmp_header = outMskNii.header
msk_lungs = resizeNii(path_lungs, newSize=outSize,
parOrder=0).get_data()
outMsk = outMskNii.get_data().astype(np.uint8)
outMsk[msk_lungs < 0.5] = 0
outMskNii = nib.Nifti1Image(outMsk.copy().astype(np.uint8),
tmp_affine,
header=tmp_header)
# if threshold is not None:
# outMskNii = nib.Nifti1Image((outMskNii.get_data() > threshold).astype(np.float16),
# outMskNii.affine,
# header=outMskNii.header)
if pathOutNii is not None:
nib.save(outMskNii, pathOutNii)
# pathOutNii = '%s-segm.nii.gz' % pathInpNii
else:
return outMskNii
def test_resize_nii(self):
self.assertTrue(os.path.isdir(self.wdir))
lstPathNii = sorted(glob.glob('%s/*.nii.gz' % self.wdir))
numNii = len(lstPathNii)
self.assertTrue((numNii > 0))
pathNii = lstPathNii[0]
newSize = (128, 128, 128)
niiResiz = preproc.resizeNii(pathNii, newSize=newSize)
self.assertTrue(niiResiz.shape == newSize)
def segmentLungs25D(pathInpNii,
dirWithModel,
pathOutNii=None,
outSize=None,
batchSize=8,
isDebug=False,
threshold=None):
if isinstance(pathInpNii, str): # or isinstance(pathInpNii,unicode):
isInpFromFile = True
if not os.path.isfile(pathInpNii):
raise Exception('Cant find input file [%s]' % pathInpNii)
else:
isInpFromFile = False
if not os.path.isdir(dirWithModel):
raise Exception('Cant find directory with model [%s]' % dirWithModel)
if pathOutNii is not None:
outDir = os.path.dirname(os.path.abspath(pathOutNii))
if not os.path.isdir(outDir):
raise Exception(
'Cant find output directory [%s], create directory for output file before this call'
% outDir)
batcherInfer = BatcherCTLung2D()
batcherInfer.loadModelForInference(pathModelJson=dirWithModel,
pathMeanData=dirWithModel)
if isDebug:
batcherInfer.model.summary()
lstPathNifti = [pathInpNii]
ret = batcherInfer.inference(lstPathNifti,
batchSize=batchSize,
isDebug=isDebug)
outMsk = ret[0]
if isInpFromFile:
tmpNii = nib.load(pathInpNii)
else:
tmpNii = pathInpNii
#
outMskNii = nib.Nifti1Image(outMsk.copy().astype(np.float16),
tmpNii.affine,
header=tmpNii.header)
# resize if need:
if outSize is not None:
outMskNii = resizeNii(outMskNii, newSize=outSize)
# threshold if need:
if threshold is not None:
outMskNii = nib.Nifti1Image(
(outMskNii.get_data() > threshold).astype(np.float16),
outMskNii.affine,
header=outMskNii.header)
# save if output path is present
if pathOutNii is not None:
nib.save(outMskNii, pathOutNii)
# pathOutNii = '%s-segm.nii.gz' % pathInpNii
else:
return outMskNii
def test_lesion_segmentation(self):
niiOriginal = nib.load(self.pathInpNii)
inpSize = niiOriginal.shape
sizeProcessing = (128, 128, 64)
niiResized = resizeNii(pathNii=niiOriginal, newSize=sizeProcessing)
niiMask = segmentLesions3D(niiResized,
dirWithModel=self.pathModelLesion,
pathOutNii=None,
outSize=inpSize,
threshold=None)
foutMskLesion = '%s-msk-lesion-test.nii.gz' % self.pathInpNii
nib.save(niiMask, foutMskLesion)
self.assertTrue(inpSize == niiMask.shape)
def test_lung_segmentation(self):
#
niiOriginal = nib.load(self.pathInpNii)
inpSize = niiOriginal.shape
sizeProcessing = (256, 256, 64)
niiResized = resizeNii(pathNii=niiOriginal, newSize=sizeProcessing)
niiMask = segmentLungs25D(niiResized,
dirWithModel=self.pathModelLung,
pathOutNii=None,
outSize=inpSize,
threshold=0.5)
foutMskLung = '%s-msk-lung-test.nii.gz' % self.pathInpNii
nib.save(niiMask, foutMskLung)
self.assertTrue(inpSize == niiMask.shape)
default=None,
required=True,
help='path to index with images')
args = parser.parse_args()
#
path_idx = args.path_idx
shape_4lung = (256, 256, 64)
wdir = os.path.dirname(path_idx)
data_csv = pd.read_csv(path_idx)
paths_img = [os.path.join(wdir, xx) for xx in data_csv['path_img']]
num_img = len(paths_img)
#
for ii, path_nii in enumerate(paths_img):
print(' [{}/{}] * {}'.format(ii, num_img, paths_img))
t1 = time.time()
dataNii = nib.load(path_nii)
shape_orig = dataNii.shape
nii_resiz_4lung = resizeNii(dataNii, shape_4lung)
lungMask = segmentLungs25D(
nii_resiz_4lung,
dirWithModel=path_model_lung,
pathOutNii=None,
outSize=shape_orig,
# outSize=shape4Lung,
threshold=0.5)
dt = time.time() - t1
path_out_lungs = '{}-lungs.nii.gz'.format(path_nii)
print('\t\tdone, dt={:0.2f} (s), export result to [{}]'.format(
dt, path_out_lungs))
nib.save(lungMask, path_out_lungs)
def api_segmentLungAndLesion(dirModelLung,
dirModelLesion,
series,
ptrLogger=None,
shape4Lung=(256, 256, 64),
shape4Lesi=(512, 512, 256),
gpuMemUsage=0.4):
# (1) msg-helpers
def msgInfo(msg):
if ptrLogger is not None:
ptrLogger.info(msg)
else:
print(msg)
def msgErr(msg):
if ptrLogger is not None:
ptrLogger.error(msg)
else:
print(msg)
# (2.1) check data
if not series.isInitialized():
msgErr('Series is not initialized, skip .. [{0}]'.format(series))
return False
# if not series.isDownloaded():
# msgErr('Series data is not downloaded, skip .. [{0}]'.format(series))
# return False
if not series.isConverted():
msgErr(
'Series DICOM data is not converted to Nifti format, skip .. [{0}]'
.format(series))
return False
# (2.2) check existing files
pathNii = series.pathConvertedNifti(isRelative=False)
# pathSegmLungs = series.pathPostprocLungs(isRelative=False)
pathSegmLungs = series.pathPostprocLungs(isRelative=False)
pathSegmLesions = series.pathPostprocLesions2(isRelative=False)
if os.path.isfile(pathSegmLungs) and os.path.isfile(pathSegmLesions):
msgInfo('Series data is already segmented, skip task ... [{0}]'.format(
series))
return False
else:
# (2.3.0) TF GPU memory usage constraints
# FIXME:temporary fix, in future: apped memory usage parameter in application config
import tensorflow as tf
import keras.backend as K
from keras.backend.tensorflow_backend import set_session
if K.image_dim_ordering() == 'tf':
config = tf.ConfigProto()
config.gpu_options.per_process_gpu_memory_fraction = gpuMemUsage
set_session(tf.Session(config=config))
# (2.3.1) load and resize
try:
dataNii = nib.load(pathNii)
shapeOrig = dataNii.shape
niiResiz4Lung = resizeNii(dataNii, shape4Lung)
except Exception as err:
msgErr(
'Cant load and resize input nifti file [{0}] : {1}, for series [{2}]'
.format(pathNii, err, series))
return False
# (2.3.2) segment lungs
try:
if not os.path.isfile(pathSegmLungs):
K.clear_session()
lungMask = segmentLungs25D(
niiResiz4Lung,
dirWithModel=dirModelLung,
pathOutNii=None,
outSize=shapeOrig,
# outSize=shape4Lung,
threshold=0.5)
nib.save(lungMask, pathSegmLungs)
else:
pass
# lungMask = nib.load(pathSegmLungs)
except Exception as err:
msgErr('Cant segment lungs for file [{0}] : {1}, for series [{2}]'.
format(pathNii, err, series))
return False
# (2.3.3) segment lesions
try:
if not os.path.isfile(pathSegmLesions):
# lesionMask = segmentLesions3D(niiResiz4Lesi,
# dirWithModel=dirModelLesion,
# pathOutNii=None,
# outSize=shapeOrig,
# # outSize=shape4Lung,
# threshold=None)
K.clear_session()
shape_lesions = [512, 512, 256]
nii_lung_resiz4lesion = resizeNii(pathSegmLungs,
shape_lesions,
parOrder=0)
nii_data_resiz4lesion = resizeNii(dataNii,
shape_lesions,
parOrder=1)
lesionMask, lesionMaskVal = segmentLesions3Dv3(
nii_data_resiz4lesion,
dir_with_model=dirModelLesion,
nii_lings=nii_lung_resiz4lesion,
path_out_nii=None,
# outSize=shapeOrig,
# outSize=shape4Lung,
threshold=None)
# lesionMask = segmentLesions3Dv2(niiResiz4Lesi,
# dirWithModel=dirModelLesion,
# pathOutNii=None,
# outSize=shapeOrig,
# # outSize=shape4Lung,
# threshold=None,
# path_lungs=pathSegmLungs)
# (2.3.4) save results
try:
lesionMask = resizeNii(lesionMask, shapeOrig, parOrder=0)
nib.save(lesionMask, pathSegmLesions)
except Exception as err:
msgErr(
'Cant save segmentation results to file [{0}] : {1}, for series [{2}]'
.format(pathSegmLesions, err, series))
return False
except Exception as err:
msgErr(
'Cant segment lesions for file [{0}] : {1}, for series [{2}]'.
format(pathNii, err, series))
return False
return True