def __call__(self, data: Any):
if self.depth:
img, seg = create_test_image_3d(
self.height,
self.width,
self.depth,
self.num_objs,
self.rad_max,
self.rad_min,
self.noise_max,
self.num_seg_classes,
self.channel_dim,
self.random_state,
)
else:
img, seg = create_test_image_2d(
self.height,
self.width,
self.num_objs,
self.rad_max,
self.rad_min,
self.noise_max,
self.num_seg_classes,
self.channel_dim,
self.random_state,
)
return img, seg
def setUp(self):
im, msk = create_test_image_3d(self.im_shape[0], self.im_shape[1],
self.im_shape[2], 4, 20, 0,
self.num_classes)
self.imt = im[None, None]
self.seg1 = (msk[None, None] > 0).astype(np.float32)
self.segn = msk[None, None]
def run_test(batch_size=2, device=torch.device("cpu:0")):
im, seg = create_test_image_3d(25,
28,
63,
rad_max=10,
noise_max=1,
num_objs=4,
num_seg_classes=1)
input_shape = im.shape
img_name = make_nifti_image(im)
seg_name = make_nifti_image(seg)
ds = NiftiDataset([img_name], [seg_name],
transform=AddChannel(),
seg_transform=AddChannel(),
image_only=False)
loader = DataLoader(ds, batch_size=1, pin_memory=torch.cuda.is_available())
net = UNet(
dimensions=3,
in_channels=1,
num_classes=1,
channels=(4, 8, 16, 32),
strides=(2, 2, 2),
num_res_units=2,
)
roi_size = (16, 32, 48)
sw_batch_size = batch_size
def _sliding_window_processor(_engine, batch):
net.eval()
img, seg, meta_data = batch
with torch.no_grad():
seg_probs = sliding_window_inference(img, roi_size, sw_batch_size,
lambda x: net(x)[0], device)
return predict_segmentation(seg_probs)
infer_engine = Engine(_sliding_window_processor)
with tempfile.TemporaryDirectory() as temp_dir:
SegmentationSaver(output_path=temp_dir,
output_ext='.nii.gz',
output_postfix='seg').attach(infer_engine)
infer_engine.run(loader)
basename = os.path.basename(img_name)[:-len('.nii.gz')]
saved_name = os.path.join(temp_dir, basename,
'{}_seg.nii.gz'.format(basename))
testing_shape = nib.load(saved_name).get_fdata().shape
if os.path.exists(img_name):
os.remove(img_name)
if os.path.exists(seg_name):
os.remove(seg_name)
return testing_shape == input_shape
from monai.handlers.stats_handler import StatsHandler
from monai.handlers.mean_dice import MeanDice
from monai.visualize import img2tensorboard
from monai.data.synthetic import create_test_image_3d
from monai.handlers.utils import stopping_fn_from_metric
# assumes the framework is found here, change as necessary
sys.path.append("..")
config.print_config()
# Create a temporary directory and 50 random image, mask paris
tempdir = tempfile.mkdtemp()
for i in range(50):
im, seg = create_test_image_3d(256, 256, 256)
n = nib.Nifti1Image(im, np.eye(4))
nib.save(n, os.path.join(tempdir, 'im%i.nii.gz' % i))
n = nib.Nifti1Image(seg, np.eye(4))
nib.save(n, os.path.join(tempdir, 'seg%i.nii.gz' % i))
images = sorted(glob(os.path.join(tempdir, 'im*.nii.gz')))
segs = sorted(glob(os.path.join(tempdir, 'seg*.nii.gz')))
# Define transforms for image and segmentation
imtrans = transforms.Compose(
[Rescale(),
AddChannel(),
UniformRandomPatch((64, 64, 64)),
from monai.networks.nets.unet import UNet
from monai.networks.utils import predict_segmentation
from monai.data.synthetic import create_test_image_3d
from monai.utils.sliding_window_inference import sliding_window_inference
from monai.handlers.stats_handler import StatsHandler
from monai.handlers.mean_dice import MeanDice
config.print_config()
logging.basicConfig(stream=sys.stdout, level=logging.INFO)
tempdir = tempfile.mkdtemp()
# tempdir = './temp'
print(
'generating synthetic data to {} (this may take a while)'.format(tempdir))
for i in range(5):
im, seg = create_test_image_3d(128, 128, 128, num_seg_classes=1)
n = nib.Nifti1Image(im, np.eye(4))
nib.save(n, os.path.join(tempdir, 'im%i.nii.gz' % i))
n = nib.Nifti1Image(seg, np.eye(4))
nib.save(n, os.path.join(tempdir, 'seg%i.nii.gz' % i))
images = sorted(glob(os.path.join(tempdir, 'im*.nii.gz')))
segs = sorted(glob(os.path.join(tempdir, 'seg*.nii.gz')))
# Define transforms for image and segmentation
imtrans = transforms.Compose([Rescale(), AddChannel()])
segtrans = transforms.Compose([AddChannel()])
ds = NiftiDataset(images,
segs,
from monai.data.nifti_reader import NiftiDataset
from monai.transforms import (AddChannel, Rescale, ToTensor, UniformRandomPatch)
from monai.handlers.stats_handler import StatsHandler
from monai.handlers.mean_dice import MeanDice
from monai.visualize import img2tensorboard
from monai.data.synthetic import create_test_image_3d
from monai.handlers.utils import stopping_fn_from_metric
monai.config.print_config()
# Create a temporary directory and 50 random image, mask paris
tempdir = tempfile.mkdtemp()
for i in range(50):
im, seg = create_test_image_3d(128, 128, 128)
n = nib.Nifti1Image(im, np.eye(4))
nib.save(n, os.path.join(tempdir, 'im%i.nii.gz' % i))
n = nib.Nifti1Image(seg, np.eye(4))
nib.save(n, os.path.join(tempdir, 'seg%i.nii.gz' % i))
images = sorted(glob(os.path.join(tempdir, 'im*.nii.gz')))
segs = sorted(glob(os.path.join(tempdir, 'seg*.nii.gz')))
# Define transforms for image and segmentation
imtrans = transforms.Compose([
Rescale(),
AddChannel(),
UniformRandomPatch((96, 96, 96)),
from monai.handlers.checkpoint_loader import CheckpointLoader
from monai.handlers.stats_handler import StatsHandler
from monai.handlers.mean_dice import MeanDice
from monai import config
config.print_config()
logging.basicConfig(stream=sys.stdout, level=logging.INFO)
tempdir = tempfile.mkdtemp()
# tempdir = './temp'
print(
'generating synthetic data to {} (this may take a while)'.format(tempdir))
for i in range(5):
im, seg = create_test_image_3d(128,
128,
128,
num_seg_classes=1,
channel_dim=-1)
n = nib.Nifti1Image(im, np.eye(4))
nib.save(n, os.path.join(tempdir, 'im%i.nii.gz' % i))
n = nib.Nifti1Image(seg, np.eye(4))
nib.save(n, os.path.join(tempdir, 'seg%i.nii.gz' % i))
images = sorted(glob(os.path.join(tempdir, 'im*.nii.gz')))
segs = sorted(glob(os.path.join(tempdir, 'seg*.nii.gz')))
val_files = [{'img': img, 'seg': seg} for img, seg in zip(images, segs)]
# Define transforms for image and segmentation
val_transforms = transforms.Compose([
from parameterized import parameterized
from monai.data.synthetic import create_test_image_2d, create_test_image_3d
from monai.transforms.utils_pytorch_numpy_unification import moveaxis
from monai.utils.module import optional_import
from monai.visualize.utils import blend_images
from tests.utils import TEST_NDARRAYS
plt, has_matplotlib = optional_import("matplotlib.pyplot")
TESTS = []
for p in TEST_NDARRAYS:
image, label = create_test_image_2d(100, 101)
TESTS.append((p(image), p(label)))
image, label = create_test_image_3d(100, 101, 102)
TESTS.append((p(image), p(label)))
@skipUnless(has_matplotlib, "Matplotlib required")
class TestBlendImages(unittest.TestCase):
@parameterized.expand(TESTS)
def test_blend(self, image, label):
blended = blend_images(image[None], label[None])
self.assertEqual(type(image), type(blended))
if isinstance(blended, torch.Tensor):
self.assertEqual(blended.device, image.device)
blended = blended.cpu().numpy()
self.assertEqual((3, ) + image.shape, blended.shape)
blended = moveaxis(blended, 0, -1) # move RGB component to end
from tests.utils import TEST_NDARRAYS
plt, has_matplotlib = optional_import("matplotlib.pyplot")
def get_alpha(img):
return 0.5 * np.arange(img.size).reshape(img.shape) / img.size
TESTS = []
for p in TEST_NDARRAYS:
image, label = create_test_image_2d(100, 101, channel_dim=0)
TESTS.append((p(image), p(label), 0.5))
TESTS.append((p(image), p(label), p(get_alpha(image))))
image, label = create_test_image_3d(100, 101, 102, channel_dim=0)
TESTS.append((p(image), p(label), 0.5))
TESTS.append((p(image), p(label), p(get_alpha(image))))
@skipUnless(has_matplotlib, "Matplotlib required")
class TestBlendImages(unittest.TestCase):
@parameterized.expand(TESTS)
def test_blend(self, image, label, alpha):
blended = blend_images(image, label, alpha)
self.assertEqual(type(image), type(blended))
if isinstance(blended, torch.Tensor):
self.assertEqual(blended.device, image.device)
blended = blended.cpu().numpy()
self.assertEqual((3,) + image[0].shape, blended.shape)