質問編集履歴
1
コードの変更
test
CHANGED
|
File without changes
|
test
CHANGED
|
@@ -34,109 +34,7 @@
|
|
|
34
34
|
|
|
35
35
|
```python
|
|
36
36
|
|
|
37
|
-
|
|
37
|
+
|
|
38
|
-
|
|
39
|
-
#mhd読み込みtrain
|
|
40
|
-
|
|
41
|
-
def load_itk(filename):
|
|
42
|
-
|
|
43
|
-
imglist = []
|
|
44
|
-
|
|
45
|
-
spacinglist = []
|
|
46
|
-
|
|
47
|
-
|
|
48
|
-
|
|
49
|
-
for i in range(len(train_a_FILE)):
|
|
50
|
-
|
|
51
|
-
itkimage = sitk.ReadImage(filename)
|
|
52
|
-
|
|
53
|
-
raw_image = sitk.GetArrayFromImage(itkimage)
|
|
54
|
-
|
|
55
|
-
raw_array = np.asarray(raw_image, dtype=np.float32)
|
|
56
|
-
|
|
57
|
-
# raw_array = raw_array.transpose(1, 2, 0)
|
|
58
|
-
|
|
59
|
-
|
|
60
|
-
|
|
61
|
-
imglist.append(raw_array)
|
|
62
|
-
|
|
63
|
-
|
|
64
|
-
|
|
65
|
-
raw_spacing = itkimage.GetSpacing()
|
|
66
|
-
|
|
67
|
-
spacinglist.append(raw_spacing)
|
|
68
|
-
|
|
69
|
-
imgdatas = np.asarray(imglist)
|
|
70
|
-
|
|
71
|
-
return raw_array , imgdatas , spacinglist
|
|
72
|
-
|
|
73
|
-
|
|
74
|
-
|
|
75
|
-
|
|
76
|
-
|
|
77
|
-
#%%
|
|
78
|
-
|
|
79
|
-
###train
|
|
80
|
-
|
|
81
|
-
train_a_datalist = []
|
|
82
|
-
|
|
83
|
-
train_b_datalist=[]
|
|
84
|
-
|
|
85
|
-
for i in range(len(train_FILE)):
|
|
86
|
-
|
|
87
|
-
a_train_image,aimagedatas,a_train_spacinglist=load_itk(train_a_FILE[i])
|
|
88
|
-
|
|
89
|
-
b_train_image,bimagedatas,b_train_spacinglist=load_itk(train_b_FILE[i])
|
|
90
|
-
|
|
91
|
-
|
|
92
|
-
|
|
93
|
-
#入力チェック時のエラー：input_1が4次元であることを期待したが、形状が(48, 180, 180)の配列を得たとエラーが出るので、newaxisで次元追加
|
|
94
|
-
|
|
95
|
-
a_train_image=a_train_image[:,:,:,np.newaxis]
|
|
96
|
-
|
|
97
|
-
a_train_image=a_train_image[:,:,:,np.newaxis]
|
|
98
|
-
|
|
99
|
-
a_train_image = np.array(a_train_image, dtype=np.float32)
|
|
100
|
-
|
|
101
|
-
b_train_image = np.array(b_train_image, dtype=np.float32)
|
|
102
|
-
|
|
103
|
-
|
|
104
|
-
|
|
105
|
-
#画像の4分割したものをとる
|
|
106
|
-
|
|
107
|
-
a_len = len(a_train_image)
|
|
108
|
-
|
|
109
|
-
a_min_i = a_len//4
|
|
110
|
-
|
|
111
|
-
a_max_i= a_len - a_min_i
|
|
112
|
-
|
|
113
|
-
|
|
114
|
-
|
|
115
|
-
b_len = len(b_train_image)
|
|
116
|
-
|
|
117
|
-
b_min_i = b_len//4
|
|
118
|
-
|
|
119
|
-
b_max_i= b_len - b_min_i
|
|
120
|
-
|
|
121
|
-
|
|
122
|
-
|
|
123
|
-
|
|
124
|
-
|
|
125
|
-
train_a_slice=np.array(a_train_image[a_min_i:a_max_i,:,:])
|
|
126
|
-
|
|
127
|
-
|
|
128
|
-
|
|
129
|
-
train_a_datalist.append(train_a_slice)
|
|
130
|
-
|
|
131
|
-
|
|
132
|
-
|
|
133
|
-
|
|
134
|
-
|
|
135
|
-
train_b_slice=np.array(b_train_image[b_min_i:b_max_i,:,:])
|
|
136
|
-
|
|
137
|
-
|
|
138
|
-
|
|
139
|
-
train_b_datalist.append(train_b_slice)
|
|
140
38
|
|
|
141
39
|
|
|
142
40
|
|