ini = ini_temp()
#np.save(weights_path+r'result_test/avg_direct.npy', ini)
#import SimpleITK as sitk
#avg = sitk.ReadImage(r'/home/dkuang/LPBA40/avg152T1_brain.nii')
#avg_data = sitk.GetArrayViewFromImage(avg)
#ini = avg_data/np.max(avg_data)
ini = np.expand_dims(np.expand_dims(ini, 3),0)
#print(ini.shape)
'''
test transform
'''
ini1 = transform(np.zeros([1,res1, res2, res3, 3]), ini, (res1, res2, res3))
print(np.all(ini1==ini))
print([np.min(ini1-ini), np.max(ini1-ini)])
#alpha = 0.5
#for i in range(20):
# disp_sum = 0
# for sample in samples:
# tgt = np.load(datapath + sample)
# tgt = np.expand_dims(np.expand_dims(tgt, 3),0)
# pair = np.stack([ini, tgt], 4)[...,0]
# print(np.all(pair[0,...,1] == tgt[0,...,0]))
#print(tgt.shape)
# disp_sum+=sdn.predict(pair)
# disp_sum = disp_sum/len(samples)
label_list = list(np.unique(label1_data))[1:]
dice_before, dice_after = [], []
'''
Different ways to warp the label mask makes huge differences...
'''
brain_test1 = np.expand_dims(brain_test1, 0)
warped_brain, deformation = sdn.predict(brain_test1)
from Utils import transform
for i in label_list:
dice_before.append(Dice(label2_data == i, label1_data == i))
seg = transform(deformation,
np.expand_dims(np.expand_dims(label1_data == i, 3), 0),
(res1, res2, res3))
dice_after.append(Dice(label2_data == i, np.rint(seg) > 0))
#print("ROI {} result appended..".format(i))
count_worse = 0
count_equal = 0
count_better = 0
for i in range(56):
if dice_after[i] < dice_before[i]:
count_worse += 1
] + [101, 102, 121, 122] + [i for i in range(161, 167, 1)] + [181, 182]
dice_before = np.zeros([len(label_test), 56])
dice_after = np.zeros([len(label_test), 56])
for test_ind, test_label in enumerate(label_test):
brain_test = np.load(datapath + test_label)
brain_test_label1 = np.load(labelpath + test_label[:2] + '.npy')
brain_test_label2 = np.load(labelpath + test_label[2:4] + '.npy')
brain_test = np.expand_dims(brain_test, 0)
deformation = sdn.predict(brain_test)
for label_ind, i in enumerate(label_list):
dice_before[test_ind, label_ind] = Dice(brain_test_label2 == i,
brain_test_label1 == i)
seg = transform(
deformation,
np.expand_dims(np.expand_dims(brain_test_label1 == i, 3), 0),
(res1, res2, res3))
dice_after[test_ind, label_ind] = Dice(brain_test_label2 == i,
np.rint(seg) > 0)
#print("ROI {} result appended..".format(i))
print('Test sample {}\'s evaluation completed.'.format(test_ind + 1))
#print("writing the first to log....")
#test_log = open('result_test/SDN3dTest.txt', 'w')
#test_log.write("Brain pair: {}\n".format(label_test[0]))
#test_log.write("*"*40+"\n")
#test_log.write("Before Registration"+"\t" + "After Registration\n")
#for before, after in zip(dice_before[0], dice_after[0]):
#test_log.write("{0:.6f}".format(before)+"\t" + "{0:.6f}".format(after)+'\n')
#test_log.close()
dice_before = np.zeros([len(train_list), 56])
for test_ind, test_label in enumerate(train_list):
brain_test_mov = np.load(datapath+test_label[:2]+'.npy')
brain_test_fix = np.load(datapath+test_label[2:]+'.npy')
brain_test = np.stack([brain_test_mov, brain_test_fix], 3)
brain_test_label1 = np.load(labelpath+test_label[:2]+'.npy')
brain_test_label2 = np.load(labelpath+test_label[2:4]+'.npy')
brain_test = np.expand_dims(brain_test, 0)
warped_vol, deformation = sdn_pred.predict(brain_test)
np.save('/home/dkuang/{}.npy'.format(test_label), warped_vol[0,...,0])
for label_ind, i in enumerate(label_list):
seg = transform(deformation, np.expand_dims(np.expand_dims(brain_test_label1==i, 3),0), (res1, res2, res3))
dice_before[test_ind, label_ind] = Dice(brain_test_label2 == i, np.rint(seg)>0)
print(dice_before)
print(np.mean(dice_before, 1))
sdn_refine.compile(loss = ['mse', total_variation_loss],
loss_weights = par['loss_weights'],
optimizer = Adam(lr = par['lr'], decay = 1e-5) )
print(sdn_refine.summary())
gen_train = vol_gen_refine(datapath, predpath,train_list, par['batch_size'])
history = sdn_refine.fit_generator(gen_train, steps_per_epoch = len(train_list)/par['batch_size'], epochs = par['epochs'], use_multiprocessing = True, verbose=1)
brain_utr_data = sitk.GetArrayViewFromImage(brain_utr)
utr_warped_img = brain_utr_data / np.max(brain_utr_data)
label_utr = sitk.ReadImage(
utillabelpath + 'S{}toS{}.fsl152.structure.label.nii'.format(mov, fix))
utr_warped_label = sitk.GetArrayViewFromImage(label_utr)
brain_test = np.load(datapath + test_label)
brain_test_label1 = np.load(labelpath + '{}.npy'.format(mov))
brain_test_label2 = np.load(labelpath + '{}.npy'.format(fix))
brain_test = np.expand_dims(brain_test, 0)
deformation = sdn.predict(brain_test)
sdn_warped_img = transform(deformation,
np.expand_dims(brain_test[..., 0], 4),
(res1, res2, res3))
sse_before[ind] = np.sum((brain_test[..., 0] - brain_test[..., 1])**2)
sse_sdn[ind] = np.sum(
(brain_test[0, ..., 1] - sdn_warped_img[0, ..., 0])**2)
sse_utr[ind] = np.sum((brain_test[0, ..., 1] - utr_warped_img)**2)
#print("sse_before: {}".format(sse_before))
#print("sse_sdn: {}".format(sse_sdn))
#print("sse_utr: {}".format(sse_utr))
for label_ind, i in enumerate(label_list):
dice_before[ind, label_ind] = Dice(brain_test_label2 == i,
brain_test_label1 == i)
seg = transform(
