def process_image(name, do_closing, closing_structure):
image_train, _, _ = image_read_write.load_itk_image(name)
name = name.replace("mask", "truth")
image_truth, _, _ = image_read_write.load_itk_image(name)
truth = np.zeros(image_truth.shape, dtype=np.uint8)
truth[image_truth > 0] = 1
if do_closing:
image_train = binary_closing(image_train, closing_structure, 1)
image_train = binary_closing(image_train,
[[[1]], [[1]], [[1]], [[1]], [[1]]], 1)
score = calculate_dice(image_train, truth, name)
return score
def candidates_to_image(cands, radius):
image_names = []
for subset in xrange(0, 10):
subset_names = glob.glob("../data/subset{0}/*.mhd".format(subset))
names = [os.path.split(x)[1].replace('.mhd', '') for x in subset_names]
image_names.append(names)
previous_candidate = ""
images = []
image = []
origin = []
spacing = []
number = 0
for candidate in tqdm(cands.values):
if candidate[0] != previous_candidate:
number = 0
previous_candidate = candidate[0]
for image_subset in xrange(0, 10):
if candidate[0] in image_names[image_subset]:
image, origin, spacing = load_itk_image(
"../data/subset{0}/{1}.mhd".format(
image_subset, candidate[0]))
break
coords = world_2_voxel([candidate[3], candidate[2], candidate[1]],
origin, spacing)
im = image_part_from_candidate(image, coords, radius)
#images.append(im)
if candidate[4]:
label = "true"
else:
label = "false"
scipy.misc.imsave(
'../data/samples/{0}_{1}_{2}.jpg'.format(candidate[0], number,
label), im)
number += 1
return images
def candidates_to_image(cands,radius):
image_names = []
for subset in xrange(0,10):
subset_names = glob.glob("../data/subset{0}/*.mhd".format(subset))
names = [os.path.split(x)[1].replace('.mhd','') for x in subset_names]
image_names.append(names)
previous_candidate = ""
images = []
image = []
origin = []
spacing = []
number = 0
for candidate in tqdm(cands.values):
if candidate[0] != previous_candidate:
number = 0
previous_candidate = candidate[0]
for image_subset in xrange(0,10):
if candidate[0] in image_names[image_subset]:
image,origin,spacing = load_itk_image("../data/subset{0}/{1}.mhd".format(image_subset,candidate[0]))
break
coords = world_2_voxel([candidate[3],candidate[2],candidate[1]],origin,spacing)
im = image_part_from_candidate(image,coords,radius)
#images.append(im)
if candidate[4]:
label = "true"
else:
label = "false"
scipy.misc.imsave('../data/samples/{0}_{1}_{2}.jpg'.format(candidate[0],number,label), im)
number += 1
return images
def load_data(image_names):
print "loading images"
images = []
origins= []
spacings = []
for image in image_names:
i,o,s = image_read_write.load_itk_image(image)
images.append(i)
origins.append(o)
spacings.append(s)
print "done"
return images,origins,spacings
def process_failure(name):
name = name.replace("mask","truth")
name2 = name.replace("truth","")
image,_,_ = image_read_write.load_itk_image(name2)
#image_cropped = image[:,120:420,60:460]
image_mask = np.zeros(image.shape)
center = 256
cc,rr = circle(center+20,center,160)
image_mask[:,cc,rr] = 1
image[image>threshold]=0
image[image!=0]=1
image = image*image_mask
#image_cropped[image_cropped>threshold]=0
#image_cropped[image_cropped!=0]=1
kernel20 = np.zeros((15,15))
cc,rr = circle(7,7,8)
kernel20[cc,rr]=1
image = binary_closing(image, [kernel20],1)
#image[:,:,:]=0
#image[:,120:420,60:460]=image_cropped
truth,_,_ = image_read_write.load_itk_image(name)
print evaluator.calculate_dice(image,truth,name)
image = np.array(image,dtype=np.int8)
def process_failure(name):
name = name.replace("mask", "truth")
name2 = name.replace("truth", "")
image, _, _ = image_read_write.load_itk_image(name2)
#image_cropped = image[:,120:420,60:460]
image_mask = np.zeros(image.shape)
center = 256
cc, rr = circle(center + 20, center, 160)
image_mask[:, cc, rr] = 1
image[image > threshold] = 0
image[image != 0] = 1
image = image * image_mask
#image_cropped[image_cropped>threshold]=0
#image_cropped[image_cropped!=0]=1
kernel20 = np.zeros((15, 15))
cc, rr = circle(7, 7, 8)
kernel20[cc, rr] = 1
image = binary_closing(image, [kernel20], 1)
#image[:,:,:]=0
#image[:,120:420,60:460]=image_cropped
truth, _, _ = image_read_write.load_itk_image(name)
print evaluator.calculate_dice(image, truth, name)
image = np.array(image, dtype=np.int8)
def remove_pipeline(data_per_slice, imagename):
# print 'reading for ' + imagename
if DO_OUTSIDE_LUNG_REMOVAL:
lungmasklocation = MHDLOCATIONS + imagename + '.mhd'
numpyImage, numpyOrigin, numpySpacing = image_read_write.load_itk_image(
lungmasklocation)
to_remove = []
for coords in data_per_slice:
voxel = ca.world_2_voxel(coords[::-1], numpyOrigin, numpySpacing)
if ((numpyImage[voxel[0]][voxel[1]][voxel[2]] != 4)
and (numpyImage[voxel[0]][voxel[1]][voxel[2]] != 3)):
to_remove.append(coords)
for coords in to_remove:
data_per_slice.remove(coords)
return data_per_slice
#print np.linalg.norm(image_coord-coords)
for x in xrange(-diameter,diameter):
for y in xrange(-diameter,diameter):
for z in xrange(-diameter,diameter):
coords = candidates.world_2_voxel(np.array((coord_x+x,coord_y+y,coord_z+z)),origin,spacing)
if np.linalg.norm(image_coord-coords) < diameter:
image_mask[coords[0],coords[1],coords[2]] = 1
print np.sum(image_mask)
# for i in xrange(0,300):
# if np.sum(image_mask[i,:,:])>5:
# plt.imshow(image_mask[150,:,:])
# plt.show()
# break
return image_mask
#image_read_write.save_itk(image_mask,"../data/circle_sample.mhd")
if __name__ == "__main__":
for i in xrange(0,2):
cads = candidates.load_candidates("/data/annotations.csv")
#index = 2
image_names = glob.glob("../data/subset{}/*.mhd".format(i))
for image_name in image_names:
image,origin,spacing = image_read_write.load_itk_image(image_name)
name = os.path.split(image_name)[1].replace('.mhd','')
image_cads = cads[cads['seriesuid'] == name]
print name
image_mask=draw_circles(image,image_cads,origin,spacing)
create_lung_segmented_same_spacing_data.save_itk(image_mask,origin,spacing,"../data/annotation_masks/subset{}/{}.mhd".format(i,name))
#coords = blob.blob_image('../data/hoi.mhd')
#with open('../data/hoi_coords.pkl','w') as f:
# pickle.dump(coords, f)
with open('../data/hoi_coords.pkl','r') as f:
candidates = pickle.load(f)
coords = []
#coords = [y for y in [x for x in candidates]]
#slice, blob, xyz
for slice in candidates:
#print slice
for blob in slice:
coords.append(blob)
#print coords
image, origin, spacing = load_itk_image('../data/hoi.mhd')
world_coords = np.array([voxel_2_world(y,origin,spacing) for y in coords])
#print world_coords
candidates = coords_to_candidates(world_coords, '1.3.6.1.4.1.14519.5.2.1.6279.6001.105756658031515062000744821260')
print len(candidates)
candidates = merge_candidates(candidates)
print len(candidates)
save_candidates('../data/hoi_candidates.csv', candidates)
import numpy as np
import image_read_write
import matplotlib.pyplot as plt
#orientation can be: x,y,z
def show_slice(image,coordinate,orientation):
if orientation == "x":
return image[:,coordinate,:]
elif orientation == "y":
return image[:,:,coordinate]
else:
return image[coordinate,:,:]
if __name__ == "__main__":
#image,_,_= image_read_write.load_itk_image("../data/subset0/subset0/1.3.6.1.4.1.14519.5.2.1.6279.6001.105756658031515062000744821260.mhd")
image,_,_= image_read_write.load_itk_image("../data/subset0/hoi.mhd")
print image.shape
show_slice(image,150,"z")
# pickle.dump(coords, f)
with open('../data/hoi_coords.pkl', 'r') as f:
candidates = pickle.load(f)
coords = []
#coords = [y for y in [x for x in candidates]]
#slice, blob, xyz
for slice in candidates:
#print slice
for blob in slice:
coords.append(blob)
#print coords
image, origin, spacing = load_itk_image('../data/hoi.mhd')
world_coords = np.array(
[voxel_2_world(y, origin, spacing) for y in coords])
#print world_coords
candidates = coords_to_candidates(
world_coords,
'1.3.6.1.4.1.14519.5.2.1.6279.6001.105756658031515062000744821260')
print len(candidates)
candidates = merge_candidates(candidates)
print len(candidates)
save_candidates('../data/hoi_candidates.csv', candidates)
def process_image(name):
image,_,_ = image_read_write.load_itk_image(name)
volume = np.sum(image)/np.product(image.shape)
return volume
def process_image(name):
image, _, _ = image_read_write.load_itk_image(name)
volume = np.sum(image) / np.product(image.shape)
return volume
np.array((coord_x + x, coord_y + y, coord_z + z)),
origin, spacing)
if np.linalg.norm(image_coord - coords) < diameter:
image_mask[coords[0], coords[1], coords[2]] = 1
print np.sum(image_mask)
# for i in xrange(0,300):
# if np.sum(image_mask[i,:,:])>5:
# plt.imshow(image_mask[150,:,:])
# plt.show()
# break
return image_mask
#image_read_write.save_itk(image_mask,"../data/circle_sample.mhd")
if __name__ == "__main__":
for i in xrange(0, 2):
cads = candidates.load_candidates("/data/annotations.csv")
#index = 2
image_names = glob.glob("../data/subset{}/*.mhd".format(i))
for image_name in image_names:
image, origin, spacing = image_read_write.load_itk_image(
image_name)
name = os.path.split(image_name)[1].replace('.mhd', '')
image_cads = cads[cads['seriesuid'] == name]
print name
image_mask = draw_circles(image, image_cads, origin, spacing)
create_lung_segmented_same_spacing_data.save_itk(
image_mask, origin, spacing,
"../data/annotation_masks/subset{}/{}.mhd".format(i, name))
