def gaus_prefilter():
traindataset = Train_dataset(1)
XT_total = traindataset.data_true(654)
volume_real = XT_total[0][:, :, :, np.newaxis]
filename_real = os.path.join(DEFAULT_SAVE_PATH_PREDICTIONS, 'real.nii.gz')
img_volume_gen = nib.Nifti1Image(volume_real, np.eye(4))
img_volume_gen.to_filename(filename_real)
x_generator = gaussian_filter(volume_real, sigma=1)
x_generator = zoom(x_generator, [0.5, 0.5, 0.5, 1], prefilter=True)
filename_gen = os.path.join(DEFAULT_SAVE_PATH_PREDICTIONS, 'gaus_prefilter_ds2.nii.gz')
img_volume_gen = nib.Nifti1Image(x_generator, np.eye(4))
img_volume_gen.to_filename(filename_gen)
x_generator = zoom(x_generator, [0.25, 0.25, 0.25, 1], prefilter=True)
filename_gen = os.path.join(DEFAULT_SAVE_PATH_PREDICTIONS, 'gaus_prefilter_ds4.nii.gz')
img_volume_gen = nib.Nifti1Image(x_generator, np.eye(4))
img_volume_gen.to_filename(filename_gen)
def evaluate(img_width, img_height, img_depth, upsampling_factor):
# dataset & variables
traindataset = Train_dataset(1)
iterations = math.ceil(
(len(traindataset.subject_list) *
0.2)) # 817 subjects total. De 0 a 654 training. De 654 a 817 test.
print(len(traindataset.subject_list))
print(iterations)
totalpsnr = 0
totalssim = 0
array_psnr = np.empty(iterations)
array_ssim = np.empty(iterations)
batch_size = 1
div_patches = 4
num_patches = traindataset.num_patches
img_width = img_width # 224
img_height = img_height # 224
img_depth = img_depth # 152
# define model
t_input_gen = tf.placeholder('float32', [1, None, None, None, 1],
name='t_image_input_to_SRGAN_generator')
srgan_network = generator(t_input_gen,
kernel=3,
nb=6,
upscaling_factor=upsampling_factor,
is_train=False,
reuse=False,
img_width=img_width,
img_height=img_height,
img_depth=img_depth)
# restore g
sess = tf.Session(config=tf.ConfigProto(allow_soft_placement=True,
log_device_placement=False))
saver = tf.train.Saver(
tf.get_collection(tf.GraphKeys.GLOBAL_VARIABLES, scope="SRGAN_g"))
saver.restore(
sess,
tf.train.latest_checkpoint(
'/work/isanchez/g/ds4-gdl-lrdecay/subpixel'))
for i in range(0, iterations):
# extract volumes
xt_total = traindataset.data_true(
654 + i) # [[self.batch_size, 224, 224, 152]]
xt_mask = traindataset.mask(654 + i)
xg_generated = np.empty([1, 224, 224, 152, 1])
normfactor = (np.amax(xt_total[0])) / 2
x_generator = ((xt_total[0] - normfactor) / normfactor)
res = 1 / upsampling_factor
x_generator = x_generator[:, :, :, np.newaxis]
x_generator = zoom(x_generator, [res, res, res, 1])
# x_generator = gaussian_filter(x_generator, sigma=1)
xg_generated[0] = sess.run(srgan_network.outputs,
{t_input_gen: x_generator[np.newaxis, :]})
xg_generated[0] = ((xg_generated[0] + 1) * normfactor)
volume_real = xt_total[0]
volume_real = volume_real[:, :, :, np.newaxis]
volume_generated = xg_generated[0]
volume_mask = aggregate(xt_mask)
# compute metrics
max_gen = np.amax(volume_generated)
max_real = np.amax(volume_real)
if max_gen > max_real:
val_max = max_gen
else:
val_max = max_real
min_gen = np.amin(volume_generated)
min_real = np.amin(volume_real)
if min_gen < min_real:
val_min = min_gen
else:
val_min = min_real
val_psnr = psnr(np.multiply(volume_real, volume_mask),
np.multiply(volume_generated, volume_mask),
dynamic_range=val_max - val_min)
array_psnr[i] = val_psnr
totalpsnr += val_psnr
val_ssim = ssim(np.multiply(volume_real, volume_mask),
np.multiply(volume_generated, volume_mask),
dynamic_range=val_max - val_min,
multichannel=True)
array_ssim[i] = val_ssim
totalssim += val_ssim
print(val_psnr)
print(val_ssim)
# save volumes
filename_gen = os.path.join(args.path_volumes, str(i) + 'gen.nii.gz')
img_volume_gen = nib.Nifti1Image(volume_generated, np.eye(4))
img_volume_gen.to_filename(filename_gen)
filename_real = os.path.join(args.path_volumes, str(i) + 'real.nii.gz')
img_volume_real = nib.Nifti1Image(volume_real, np.eye(4))
img_volume_real.to_filename(filename_real)
print('{}{}'.format('Mean PSNR: ', array_psnr.mean()))
print('{}{}'.format('Mean SSIM: ', array_ssim.mean()))
print('{}{}'.format('Variance PSNR: ', array_psnr.var()))
print('{}{}'.format('Variance SSIM: ', array_ssim.var()))
print('{}{}'.format('Max PSNR: ', array_psnr.max()))
print('{}{}'.format('Min PSNR: ', array_psnr.min()))
print('{}{}'.format('Max SSIM: ', array_ssim.max()))
print('{}{}'.format('Min SSIM: ', array_ssim.min()))
def interpolate():
traindataset = Train_dataset(1)
iterations = math.ceil((len(traindataset.subject_list) * 0.2))
# 817 subjects total. De 0 a 654 training. De 654 a 817 test.
totalpsnr = 0
totalssim = 0
array_psnr = np.empty(iterations)
array_ssim = np.empty(iterations)
batch_size = 1
div_patches = 4
num_patches = traindataset.num_patches
img_width = 32 # 64
img_height = 32 # 64
img_depth = 23 # 46
for i in range(0, iterations):
XT_total = traindataset.data_true(654 + i)
XT_mask = traindataset.mask(654 + i)
volume_real = XT_total[0][:, :, :, np.newaxis]
#volume_real_down = zoom(gaussian_filter(volume_real, sigma=1), [0.5, 0.5, 0.5, 1], prefilter=False, order=1)
volume_real_down = zoom(volume_real, [0.5, 0.5, 0.5, 1])
volume_generated = zoom(volume_real_down, [2, 2, 2, 1])
#volume_generated = volume_generated[:, :, :, np.newaxis]
#volume_real = XT_total[0][:, :, :, np.newaxis]
volume_mask = aggregate(XT_mask)
# compute metrics
max_gen = np.amax(volume_generated)
max_real = np.amax(volume_real)
if max_gen > max_real:
max = max_gen
else:
max = max_real
min_gen = np.amin(volume_generated)
min_real = np.amin(volume_real)
if min_gen < min_real:
min = min_gen
else:
min = min_real
val_psnr = psnr(np.multiply(volume_real, volume_mask),
np.multiply(volume_generated, volume_mask),
dynamic_range=max - min)
# val_psnr = psnr(volume_real, volume_generated,
# dynamic_range=max - min)
array_psnr[i] = val_psnr
totalpsnr += val_psnr
val_ssim = ssim(np.multiply(volume_real, volume_mask),
np.multiply(volume_generated, volume_mask),
dynamic_range=max - min,
multichannel=True)
array_ssim[i] = val_ssim
totalssim += val_ssim
print(val_psnr)
print(val_ssim)
#save volumes
filename_gen = os.path.join(DEFAULT_SAVE_PATH_PREDICTIONS,
str(i) + 'gen.nii.gz')
img_volume_gen = nib.Nifti1Image(volume_generated, np.eye(4))
img_volume_gen.to_filename(filename_gen)
filename_real = os.path.join(DEFAULT_SAVE_PATH_PREDICTIONS,
str(i) + 'real.nii.gz')
img_volume_real = nib.Nifti1Image(volume_real, np.eye(4))
img_volume_real.to_filename(filename_real)
filename_down = os.path.join(DEFAULT_SAVE_PATH_PREDICTIONS,
str(i) + 'down.nii.gz')
img_volume_down = nib.Nifti1Image(volume_real_down, np.eye(4))
img_volume_down.to_filename(filename_down)
return array_psnr, array_ssim
