-
Notifications
You must be signed in to change notification settings - Fork 1
/
GetData_new.py
189 lines (144 loc) · 8.05 KB
/
GetData_new.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
# -*- coding: utf-8 -*-
"""
Created on Wed Aug 2 16:19:51 2017
@author: Think
"""
import os
import random
import numpy as np
import nibabel as nib
class GetData_new():
def __init__(self,data_dir):
self.source_list = []
self.label_dir = os.path.join(data_dir, "Labels")
self.image_dir = os.path.join(data_dir, "Images")
self.mean = 447.140490993
self.std = 303.824365028 #这是MM-WHS 2017 whole heart segmentation 的均值和方差
examples = 0
filelist = os.listdir(self.image_dir)
filelist.sort()
for file in filelist:
if not file.endswith(".nii.gz"):
continue
try:
self.source_list.append(file)
examples = examples +1
except Exception as e:
print(e)
print("finished loading images")
self.examples = examples
print("Number of examples found: ", examples)
def next_batch(self,batch_size):
random.shuffle(self.source_list)
images_list = []
labels_list = []
for i in range(batch_size):
file = self.source_list[i]
image = nib.load(os.path.join(self.image_dir, file))
label = nib.load(os.path.join(self.label_dir, file))
image_data = image.get_data()
label_data = label.get_data() #image_data and label_data : shape is (x,y,z)
#convert the image_data's shape to (x,y,z,channels)
image_x,image_y,image_z = image_data.shape
image_data = image_data[:,:,:,None]
#在进行multi-label分割的时候，应该先把label变换成[size_x,size_y,size_z,n_class]的形式，把每一种label分开
class0 = (label_data==500).astype(np.int16) #the left ventricle blood cavity
class1 = (label_data==600).astype(np.int16) # the right ventricle blood cavity
class2 = (label_data==420).astype(np.int16) # the left atrium blood cavity
class3 = (label_data==550).astype(np.int16) #the right atrium blood cavity
class4 = (label_data==205).astype(np.int16) # the myocardium of the left ventricle
class5 = (label_data==820).astype(np.int16) #the ascending aorta
class6 = (label_data==850).astype(np.int16) #the pulmonary artery
class7 = (label_data==0).astype(np.int16) #background
label_class = np.stack((class0,class1,class2,class3,class4,class5,class6,class7),axis=3)
#在分割成(64,64,64)的时候，可能会出现(64,63,64)的边界情况，需要加上判断
if image_data.shape==(64,64,64,1):
image_data = (image_data-self.mean)/self.std
images_list.append((image_data).astype(np.float32))
labels_list.append((label_class).astype(np.int16))
else:
print("the shape of input image is not (64,64,64)")
images = np.asarray(images_list)
labels = np.asarray(labels_list)
return images,labels
def next_batch_order(self,batch_size):
images_list = []
labels_list = []
for i in range(batch_size):
file = self.source_list[i]
image = nib.load(os.path.join(self.image_dir, file))
label = nib.load(os.path.join(self.label_dir, file))
image_data = image.get_data()
label_data = label.get_data() #image_data and label_data : shape is (x,y,z)
#convert the image_data's shape to (x,y,z,channels)
image_x,image_y,image_z = image_data.shape
image_data = image_data[:,:,:,None]
#在进行multi-label分割的时候，应该先把label变换成[size_x,size_y,size_z,n_class]的形式，把每一种label分开
class0 = (label_data==500).astype(np.int16)
class1 = (label_data==600).astype(np.int16)
class2 = (label_data==420).astype(np.int16)
class3 = (label_data==550).astype(np.int16)
class4 = (label_data==205).astype(np.int16)
class5 = (label_data==820).astype(np.int16)
class6 = (label_data==850).astype(np.int16)
class7 = (label_data==0).astype(np.int16)
label_class = np.stack((class0,class1,class2,class3,class4,class5,class6,class7),axis=3)
#在分割成(64,64,64)的时候，可能会出现(64,63,64)的边界情况，需要加上判断
if image_data.shape==(64,64,64,1):
image_data = (image_data-self.mean)/self.std
images_list.append((image_data).astype(np.float32))
labels_list.append((label_class).astype(np.int16))
else:
print("the shape of input image is not (64,64,64)")
images = np.asarray(images_list)
labels = np.asarray(labels_list)
return images,labels
def next_batch_order_1(self,batch_size,last_point):
images_list = []
labels_list = []
for i in range(last_point,last_point+batch_size):
file = self.source_list[i]
image = nib.load(os.path.join(self.image_dir, file))
label = nib.load(os.path.join(self.label_dir, file))
image_data = image.get_data()
label_data = label.get_data() #image_data and label_data : shape is (x,y,z)
#convert the image_data's shape to (x,y,z,channels)
image_x,image_y,image_z = image_data.shape
image_data = image_data[:,:,:,None]
#在进行multi-label分割的时候，应该先把label变换成[size_x,size_y,size_z,n_class]的形式，把每一种label分开
class0 = (label_data==500).astype(np.int16)
class1 = (label_data==600).astype(np.int16)
class2 = (label_data==420).astype(np.int16)
class3 = (label_data==550).astype(np.int16)
class4 = (label_data==205).astype(np.int16)
class5 = (label_data==820).astype(np.int16)
class6 = (label_data==850).astype(np.int16)
class7 = (label_data==0).astype(np.int16)
label_class = np.stack((class0,class1,class2,class3,class4,class5,class6,class7),axis=3)
#在分割成(64,64,64)的时候，可能会出现(64,63,64)的边界情况，需要加上判断
if image_data.shape==(64,64,64,1):
image_data = (image_data-self.mean)/self.std
images_list.append((image_data).astype(np.float32))
labels_list.append((label_class).astype(np.int16))
else:
print("the shape of input image is not (64,64,64)")
images = np.asarray(images_list)
labels = np.asarray(labels_list)
return images,labels
def next_batch_order_2(self,batch_size,last_point):
images_list = []
for i in range(last_point,last_point+batch_size):
file = self.source_list[i]
image = nib.load(os.path.join(self.image_dir, file))
image_data = image.get_data()
#convert the image_data's shape to (x,y,z,channels)
image_x,image_y,image_z = image_data.shape
image_data = image_data[:,:,:,None]
#在分割成(64,64,64)的时候，可能会出现(64,63,64)的边界情况，需要加上判断
if image_data.shape==(64,64,64,1):
image_data = (image_data-np.mean(image_data))/np.std(image_data)
images_list.append((image_data).astype(np.float32))
else:
print("the shape of input image is not (64,64,64)")
images = np.asarray(images_list)
return images