forked from kaixinjiuxing666/ADNI_nii_tensorflow
-
Notifications
You must be signed in to change notification settings - Fork 0
/
1_step_data.py
144 lines (107 loc) · 4.29 KB
/
1_step_data.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
# -*- coding: utf-8 -*-
import numpy as np
import os, random, gc, pickle
import nibabel as nib
import pandas as pd
from tqdm import tqdm
#==================== LOAD ALL IMAGES' PATH AND COMPUTE MEAN/ STD==============
#.csv是存放表格数据的格式，可用excel直接打开并编辑
metadata = pd.read_csv('/home/zhengyz/oooooo/new/data25622.csv')
totaldata = (metadata.Label != 0).values.astype('bool')
print(totaldata.sum())#合格图片总数
metaseg = pd.read_csv('/home/zhengyz/oooooo/new/seg25622.csv')
totalseg = (metaseg.Label != 0).values.astype('bool')
print(totalseg.sum())#合格图片总数
#val:192*192*160-74组:15
#tra:192*192*160-74组:59 cup_feature_guard
#val:256*256*166-36组:6
#tra:256*256*166-36组:30
#val:256*256*180-22组:4
#tra:256*256*180-22组:18
data = []
for it, im in tqdm(enumerate(metadata[totaldata].Path.values),
total=totaldata.sum(), desc='Reading MRI to memory'):
img = nib.load(im).get_data()
data.append(img)
data = np.asarray(data)
m = np.mean(data)
s = np.std(data)
print(m)
print(s)
print(data.shape)
##==================== GET NORMALIZE IMAGES====================================
X_train_input = []
X_train_target = []
X_dev_input = []
X_dev_target = []
print("Validation")
for it, im in tqdm(enumerate(metadata[:4].Path.values),
total=totaldata.sum(),desc='Reading MRI to memory'):
all_3d_data = []
img = nib.load(im).get_data()
img = (img - m) / s
img = img.astype(np.float32)
all_3d_data.append(img)
for j in range(all_3d_data[0].shape[2]):
combined_array = np.transpose(all_3d_data[0][:,:,j], (1, 0, 2))#.tolist()
combined_array.astype(np.float32)
X_dev_input.append(combined_array)
del all_3d_data
gc.collect()
for it, im in tqdm(enumerate(metaseg[:4].Path.values),
total=totalseg.sum(),desc='Reading LABEL to memory'):
all_3d_seg = []
seg_img = nib.load(im).get_data()
seg_img = seg_img.astype(np.float32)
all_3d_seg.append(seg_img)
for j in range(all_3d_seg[0].shape[2]):
combined_array = np.transpose(all_3d_seg[0][:,:,j], (1, 0, 2))#.tolist()
combined_array.astype(int)
X_dev_target.append(combined_array)
del all_3d_seg
gc.collect()
X_dev_input = np.asarray(X_dev_input, dtype=np.float32)
X_dev_target = np.asarray(X_dev_target)#, dtype=np.float32)
X_dev_target = np.minimum(X_dev_target,1)
print(X_dev_input.shape)
print(X_dev_target.shape)
print("Train")
for it, im in tqdm(enumerate(metadata[4:].Path.values),
total=totaldata.sum(),desc='Reading MRI to memory'):
all_3d_data = []
img = nib.load(im).get_data()
img = (img - m) / s
img = img.astype(np.float32)
all_3d_data.append(img)
for j in range(all_3d_data[0].shape[2]):
combined_array = np.transpose(all_3d_data[0][:,:,j], (1, 0, 2))#.tolist()
combined_array.astype(np.float32)
X_train_input.append(combined_array)
del all_3d_data
gc.collect()
for it, im in tqdm(enumerate(metaseg[4:].Path.values),
total=totalseg.sum(),desc='Reading LABEL to memory'):
all_3d_seg = []
seg_img = nib.load(im).get_data()
seg_img = seg_img.astype(np.float32)
all_3d_seg.append(seg_img)
for j in range(all_3d_seg[0].shape[2]):
combined_array = np.transpose(all_3d_seg[0][:,:,j], (1, 0, 2))#.tolist()
combined_array.astype(int)
X_train_target.append(combined_array)
del all_3d_seg
gc.collect()
X_train_input = np.asarray(X_train_input, dtype=np.float32)
X_train_target = np.asarray(X_train_target)#, dtype=np.float32)
X_train_target = np.minimum(X_train_target,1)
print(X_train_input.shape)
print(X_train_target.shape)
np.savez("256save22.npz",X_train_input,X_train_target,X_dev_input,X_dev_target)
#r=np.load("testsave.npz")
#rr=r["arr_0"]
#print(rr)
#save_dir='C:/Users/Dell/Desktop/ccc/'
#with open(save_dir + 'train_input.pickle', 'wb') as f:
# pickle.dump(X_train_input, f, protocol=4)
#with open(save_dir + 'train_target.pickle', 'wb') as f:
# pickle.dump(X_train_target, f, protocol=4)