/
dataloader.py
228 lines (144 loc) · 7.01 KB
/
dataloader.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
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
import pandas as pd
import numpy as np
import SimpleITK as sitk
import torch.utils.data as data_utils
import torch
from monai import transforms
from tqdm import tqdm
from monai.data import DataLoader
from torch.utils.data import Dataset
import os
class GetPatch(object):
def __init__(self,patchsize):
self.patchsize=patchsize
def __call__(self,image,mask=None):
if mask is None:
return self._patch(image)
else:
return self._patch(image),self._patch(mask)
def _patch(self,src):
n=self.patchsize
cnt = 0
imglist=[]
h,w,d=src.shape
maxiter=np.max([h,w,d]) // n
# print(maxiter)
for i in range(maxiter):
for j in range(maxiter):
for k in range(maxiter):
imglist.append(src[i*n:(i+1)*n,j*n:(j+1)*n,k*n:(k+1)*n][None,...])
# print(len(imglist),imglist[0].shape)
return torch.cat(imglist,0)
class Dataset(Dataset):
def __init__(self ,mode='train',transformer=False):
datacsv=pd.read_csv('/data1/dataset/splitdataTrainvalidTest.csv')
mode2mod={'train':'isTraindata','valid':'isValiddata','test':'isTestdata'}
self.mode=mode
self.KEYS=('image','mask')
mod=mode2mod[self.mode]
if self.mode=='train':
self.plug_names = datacsv[(datacsv['discard']==False) & (datacsv[mod]==True)]['filename'].to_list()
self.plug_label_map = datacsv[(datacsv['discard']==False) & (datacsv[mod]==True)]['label'].to_list()
else:
self.plug_names = datacsv[(datacsv['discard']==False) & (datacsv[mod]==True)]['filename'].to_list()
self.plug_label_map = datacsv[(datacsv['discard']==False) & (datacsv[mod]==True)]['label'].to_list()
self.label_encoder={
'0':0,
'1':1,
'2':1,
'3':1,
'4':1
}
self.transformer=transformer
self.getpatch=GetPatch(32)
def __len__(self):
return len(self.plug_names)
def _transform(self, data):
if self.mode!='train' or self.transformer:
transform=transforms.Compose([
# LoadImageD(KEYS),
transforms.EnsureChannelFirstD("image"),
transforms.AddChannelD("mask"),
# OrientationD(KEYS, axcodes='RAS'),
# SpacingD(KEYS, pixdim=(1., 1., 1.), mode=('bilinear', 'nearest')),
transforms.ScaleIntensityD(keys="image"),
# ResizeD(KEYS, (64, 64, 32), mode=('trilinear', 'nearest')),
# RandAffineD(KEYS, spatial_size=(-1, -1, -1),
# rotate_range=(0, 0, np.pi/2),
# scale_range=(0.1, 0.1),
# mode=('bilinear', 'nearest'),
# prob=1.0),
transforms.NormalizeIntensityD(keys='image'),
transforms.ToTensorD(self.KEYS),
])
else:
transform = transforms.Compose([
# transforms.LoadImageD(KEYS),
transforms.EnsureChannelFirstD("image"),
transforms.AddChannelD("mask"),
transforms.ScaleIntensityD(keys="image"),
transforms.OneOf([transforms.RandSpatialCropD(self.KEYS,roi_size=(80,80,80)),
# ResizeD(KEYS, (64, 64, 32), mode=('trilinear', 'nearest')),
transforms.RandAffineD( self.KEYS,spatial_size=(-1, -1, -1),
translate_range=(32, 32, 32),
# translate_range=(16, 16, 16),
# translate_range=(8, 8, 8),
rotate_range=(0, np.pi/4, 0),
# rotate_range=(0, np.pi/12, 0),
scale_range=(0.1, 0.1,0.1),
mode=('bilinear', 'nearest'),
prob=1.0)],[0.5,0.5]),
transforms.RandAxisFlipD(self.KEYS,prob=0.3),
transforms.NormalizeIntensityD(keys='image'),
transforms.ResizeD(self.KEYS,(128,128,128), mode=('trilinear', 'nearest')),
transforms.ToTensorD(self.KEYS),
])
return transform(data)
def getfiles(self):
return self.plug_names
def __getitem__(self, index):
plug_name = self.plug_names[index]
plug_name=os.path.join('/data1/dataset/ct_box_3d_128_clip_npy/',plug_name.split('/')[-1])
if self.mode=='train':
maskpath=os.path.join('/data1/dataset/masks',plug_name.split('/')[-1])
else:
maskpath=plug_name
dict_loader = transforms.LoadImageD(keys=("image", "mask"))
data = dict_loader({"image": plug_name,
"mask": maskpath})
data=self._transform(data)
label = self.label_encoder[str(self.plug_label_map[index])]
# image_array=np.load(plug_name,allow_pickle=True)
# img=image_array
# img=self._transform(image_array)
# label = self.label_encoder[str(self.plug_label_map[index])]
return self.getpatch(data['image'][0])[:,None,...],self.getpatch(data['mask'][0])[:,None,...],torch.tensor(label,dtype=torch.long)
if __name__=='__main__':
mytraindataset=Dataset( 'train' )
myvaliddataset=Dataset( 'valid' )
mytestdataset=Dataset( 'test' )
train_loader = DataLoader(mytraindataset, batch_size=1, shuffle=True, num_workers=1)
valid_loader = DataLoader(myvaliddataset, batch_size=1, shuffle=False, num_workers=1)
test_loader = DataLoader(mytestdataset, batch_size=1, shuffle=False, num_workers=1)
for i,m,l in mytraindataset:
# i=torch.reshape(i,(128,128,128))
print(i.shape,l,m.shape)
# showplt(i,l)
# showAsrc(i,l,A)
print(i.min(),i.max())
break
for i,_,l in myvaliddataset:
print(i.shape,l)
print(i.min(),i.max())
break
for i,l in mytestdataset:
print(i.shape,l)
print(i.min(),i.max())
break
print(len(mytraindataset),len(myvaliddataset),len(mytestdataset))
# for i,m,l in train_loader:
# print(i.shape,l,m.shape)
# # showplt(i,l)
# # showAsrc(i,l,A)
# print(i.min(),i.max())
# break