-
Notifications
You must be signed in to change notification settings - Fork 4
/
Copy pathTrain.py
202 lines (163 loc) · 7.29 KB
/
Train.py
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
# python imports
import os
import glob
import warnings
# external imports
import torch
import numpy as np
import SimpleITK as sitk
from torch.optim import Adam
import torch.utils.data as Data
# internal imports
from utils import losses
from utils.config import args
from utils.datagenerators_atlas import Dataset
from Models.STN import SpatialTransformer
from natsort import natsorted
from Models.TransMatch import TransMatch
def count_parameters(model):
model_parameters = filter(lambda p: p.requires_grad, model.parameters())
params = sum([np.prod(p.size()) for p in model_parameters])
return params
def make_dirs():
if not os.path.exists(args.model_dir):
os.makedirs(args.model_dir)
if not os.path.exists(args.log_dir):
os.makedirs(args.log_dir)
if not os.path.exists(args.result_dir):
os.makedirs(args.result_dir)
def save_image(img, ref_img, name):
img = sitk.GetImageFromArray(img[0, 0, ...].cpu().detach().numpy())
img.SetOrigin(ref_img.GetOrigin())
img.SetDirection(ref_img.GetDirection())
img.SetSpacing(ref_img.GetSpacing())
sitk.WriteImage(img, os.path.join(args.result_dir, name))
def compute_label_dice(gt, pred):
# 需要计算的标签类别，不包括背景和图像中不存在的区域
cls_lst = [21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 41, 42, 43, 44, 45, 46, 47, 48, 49, 50, 61, 62,
63, 64, 65, 66, 67, 68, 81, 82, 83, 84, 85, 86, 87, 88, 89, 90, 91, 92, 101, 102, 121, 122, 161, 162,
163, 164, 165, 166]
# cls_lst = [182]
dice_lst = []
for cls in cls_lst:
dice = losses.DSC(gt == cls, pred == cls)
dice_lst.append(dice)
return np.mean(dice_lst)
def train():
make_dirs()
device = torch.device('cuda:{}'.format(args.gpu) if torch.cuda.is_available() else 'cpu')
# 日志文件
log_name = str(args.n_iter) + "_" + str(args.lr) + "_" + str(args.alpha)
print("log_name: ", log_name)
f = open(os.path.join(args.log_dir, log_name + ".txt"), "w")
# 读入fixed图像 [D, W, H] = 160×192×160
f_img = sitk.ReadImage(args.atlas_file)
input_fixed = sitk.GetArrayFromImage(f_img)[np.newaxis, np.newaxis, ...]
vol_size = input_fixed.shape[2:]
# [B, C, D, W, H]
input_fixed_eval = torch.from_numpy(input_fixed).to(device).float()
input_fixed = np.repeat(input_fixed, args.batch_size, axis=0)
input_fixed = torch.from_numpy(input_fixed).to(device).float()
fixed_label = sitk.GetArrayFromImage(sitk.ReadImage(os.path.join(args.label_dir, "S01.delineation.structure.label.nii.gz")))[np.newaxis, np.newaxis, ...]
fixed_label = torch.from_numpy(fixed_label).to(device).float()
# 创建配准网络（net）和STN
net = TransMatch(args).to(device)
iterEpoch = 1
contTrain = False
if contTrain:
checkpoint = torch.load('./Checkpoint/500.pth')
net.load_state_dict(checkpoint)
iterEpoch = 501
STN = SpatialTransformer(vol_size).to(device)
STN_label = SpatialTransformer(vol_size, mode="nearest").to(device)
# UNet.train()
net.train()
STN.train()
opt = Adam(net.parameters(), lr=args.lr, weight_decay=0, amsgrad=True)
sim_loss_fn = losses.ncc_loss if args.sim_loss == "ncc" else losses.mse_loss
grad_loss_fn = losses.gradient_loss
# Get all the names of the training data
train_files = glob.glob(os.path.join(args.train_dir, '*.nii.gz'))
DS = Dataset(files=train_files)
print("Number of training images: ", len(DS))
DL = Data.DataLoader(DS, batch_size=args.batch_size, shuffle=True, num_workers=0, drop_last=True)
# Training loop.
for i in range(iterEpoch, args.n_iter + 1):
# Generate the moving images and convert them to tensors.
net.train()
STN.train()
print('epoch:', i)
input_moving_all = iter(DL)
for input_moving, fig_name in input_moving_all:
# [B, C, D, W, H]
fig_name = fig_name[0]
input_moving = input_moving.to(device).float()
# Run the data through the model to produce warp and flow field
flow_m2f = net(input_fixed, input_moving)
m2f = STN(input_fixed, flow_m2f)
# Calculate loss
sim_loss = sim_loss_fn(m2f, input_moving)
grad_loss = grad_loss_fn(flow_m2f)
# zero_loss = zero_loss_fn(flow_m2f, zero)
loss = sim_loss + args.alpha * grad_loss # + zero_loss
print("%d, %s, %f, %f, %f" % (i, fig_name, loss.item(), sim_loss.item(), grad_loss.item()), file=f)
# Backwards and optimize
opt.zero_grad()
loss.backward()
opt.step()
# inverse fixed image and moving image
flow_m2f = net(input_moving, input_fixed)
m2f = STN(input_moving, flow_m2f)
# Calculate loss
sim_loss = sim_loss_fn(m2f, input_fixed)
grad_loss = grad_loss_fn(flow_m2f)
# zero_loss = zero_loss_fn(flow_m2f, zero)
loss = sim_loss + args.alpha * grad_loss # + zero_loss
print("%d, %s, %f, %f, %f" % (i, fig_name, loss.item(), sim_loss.item(), grad_loss.item()), file=f)
# Backwards and optimize
opt.zero_grad()
loss.backward()
opt.step()
test_file_lst = glob.glob(os.path.join(args.test_dir, "*.nii.gz"))
net.eval()
STN.eval()
STN_label.eval()
DSC = []
for file in test_file_lst:
fig_name = file[58:60]
name = os.path.split(file)[1]
# 读入moving图像
input_moving = sitk.GetArrayFromImage(sitk.ReadImage(file))[np.newaxis, np.newaxis, ...]
input_moving = torch.from_numpy(input_moving).to(device).float()
# 读入moving图像对应的label
label_file = glob.glob(os.path.join(args.label_dir, name[:3] + "*"))[0]
input_label = sitk.GetArrayFromImage(sitk.ReadImage(label_file))
# 获得配准后的图像和label
pred_flow = net(input_fixed_eval, input_moving)
pred_img = STN(input_fixed_eval, pred_flow)
pred_label = STN_label(fixed_label, pred_flow)
# pred_label = input_label # 用于测试初始的dice值
# 计算DSC
dice = compute_label_dice(input_label, pred_label[0, 0, ...].cpu().detach().numpy())
print("{0}" .format(dice))
DSC.append(dice)
del pred_flow, pred_img, pred_label, input_moving
print(np.mean(DSC), np.std(DSC))
save_checkpoint({
'epoch': i+1,
'state_dict': net.state_dict(),
'optimizer': opt.state_dict(),
}, save_dir='experiments/1212firstrunorigincode/', filename='dsc{:.4f}epoch{:0>3d}.pth.tar'.format(np.mean(DSC), i+1))
f.close()
def save_checkpoint(state, save_dir='models', filename='checkpoint.pth.tar', max_model_num=8):
model_lists = natsorted(glob.glob(save_dir+ '*'))
while len(model_lists) > max_model_num:
os.remove(model_lists[0])
model_lists = natsorted(glob.glob(save_dir + '*'))
if not os.path.exists(save_dir):
os.makedirs(save_dir)
torch.save(state, save_dir+filename)
if __name__ == "__main__":
with warnings.catch_warnings():
warnings.filterwarnings("ignore", category=DeprecationWarning)
train()