-
Notifications
You must be signed in to change notification settings - Fork 1
/
Copy pathpre_data.py
196 lines (161 loc) · 7.9 KB
/
pre_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
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
import numpy as np
import os
import pandas as pd
import cv2
import SimpleITK as sitk
from visual_dicoms import sitk_read_dcmseries, norm_img
import matplotlib.pyplot as plt
from skimage import transform, exposure, measure
from PIL import Image
meta_path = './lsy_workspace'
data_path = 'data/data_converted/'
newimg_path = 'data/data_lsy/imgs'
newseg_path = 'data/data_lsy/masks'
newpngs_path = 'data/data_lsy/pngs'
train_data = pd.read_csv('lsy_workspace/train_8_1.csv')
val_data = pd.read_csv('lsy_workspace/validation_8_1.csv')
exter1_data = pd.read_csv('lsy_workspace/exter1.csv')
exter2_data = pd.read_csv('lsy_workspace/exter2.csv')
# if not os.path.exists('data/data_lsy/jmzxyy'):
# # os.mkdir创建一个，os.makedirs可以创建路径上多个
# os.makedirs('data/data_lsy/jmzxyy')
#########所有医院数据进行数据清洗
hosp_name = exter2_data['name'].str.split("_",expand=True)[0]
patient_path = data_path + hosp_name + '/' + exter2_data['name']
for info in patient_path:
if 'jmzxyy' in info or 'dpyy' in info:
continue
if 'zhengxianling' in info:
ori_path = r'E:\LNM_Pred\data\data_original\nbfy\zhengxianling'
seg_path = r'E:\LNM_Pred\data\data_original\nbfy\zhengxianling\zhengxianling.nii.gz'
dicom_imgs = sitk_read_dcmseries(ori_path)
segs = sitk.ReadImage(seg_path)
segs = sitk.GetArrayFromImage(segs)
for slice in range(len(segs)):
if segs[slice].any()==True:
img = dicom_imgs[slice]
seg = segs[slice]
if seg.shape != (512, 512):
img = img[-512:,:]
seg = seg[-512:,:]
window=[-1024, 1023]
img[np.where(img < window[0])] = window[0]
img[np.where(img > window[1])] = window[1]
img -= window[0]
print('image_max is ',img.max(),'min is', img.min())
hosp_patient_name = info.split('/')[3] + '.npy'
np.save(os.path.join(newimg_path, hosp_patient_name), img)
np.save(os.path.join(newseg_path, hosp_patient_name), seg)
plt.imsave(os.path.join(newpngs_path, info.split('/')[3] +'.png'), img, cmap = 'gray')
continue
######筛选出ROI最大的一张seg
if len(os.listdir(info)) != 4:
seg_files = [seg_info for seg_info in os.listdir(info) if 'mask' in seg_info]
one_oreas = []
for segfile in seg_files:
segfile_path = info + '/' + segfile
seg_info = cv2.imread(segfile_path, cv2.IMREAD_GRAYSCALE)
one_orea = np.count_nonzero(seg_info)
one_oreas.append(one_orea)
mask_path = seg_files[one_oreas.index(max(one_oreas))]
max_num = mask_path.split("_", -1)[3][:3]
img_names = [img_info for img_info in os.listdir(info) if 'img' in img_info and 'tiff'in img_info]
assert len(seg_files) == len(img_names), 'nums not match'
img_path = [img_info for img_info in os.listdir(info) if 'img' in img_info and 'tiff'in img_info and max_num in img_info][0]
######保存所有seg CT用于训练
# seg_files = [seg_info for seg_info in os.listdir(info) if 'mask' in seg_info]
# ori_files = [img_info for img_info in os.listdir(info) if 'img' in img_info and 'tiff' in img_info]
# assert len(seg_files)==len(ori_files),'seg and image is not matched'
else:
mask_path = [seg_info for seg_info in os.listdir(info) if 'mask' in seg_info][0]
img_path = [img_info for img_info in os.listdir(info) if 'img' in img_info and 'tiff' in img_info][0]
mask_path = os.path.join(info, mask_path)
img_path = os.path.join(info, img_path)
# tif = TIFF.open(img_path, mode='r')
# image = tif.read_image()
image = Image.open(img_path)
image = np.array(image, dtype='float32')
print('image_max is ',image.max(),'min is', image.min())
# MIN_BOUND = image.min()
# MAX_BOUND = image.max()
# image = (image - MIN_BOUND) / (MAX_BOUND - MIN_BOUND)
mask = cv2.imread(mask_path, cv2.IMREAD_GRAYSCALE)
mask[mask > 0] = 1
assert mask.shape == image.shape , 'shape not match'
hosp_patient_name = info.split('/')[3] + '.npy'
np.save(os.path.join(newimg_path, hosp_patient_name), image)
np.save(os.path.join(newseg_path, hosp_patient_name), mask)
plt.imsave(os.path.join(newpngs_path, info.split('/')[3] +'.png'), image, cmap = 'gray')
print(info, 'finish')
def norm_img(image): # 归一化像素值到（0，255）之间，且将溢出值取边界值
MIN_BOUND = image.min()
MAX_BOUND = image.max()
image = (image - MIN_BOUND) / (MAX_BOUND - MIN_BOUND)
image[image > 255] = 255.
image[image < 0] = 0.
return image
######获取江门中心医院 大坪医院 分割
#####################jmzxyy
base_path = 'data/data_original/jmzxyy'
seg_base_path = 'data/data_original/jmzxyy_seg'
patients_train = train_data[train_data['hospital']=='jmzxyy']
patients_val = val_data[val_data['hospital']=='jmzxyy']
patients_info = pd.concat([patients_train, patients_val], axis=0)
for index, patient_info in patients_info.iterrows():
####大坪医院多个'V'文件夹
ori_path = os.path.join(base_path, patient_info['patient'])
seg_path = [os.path.join(seg_base_path, seg_info) for seg_info in os.listdir(seg_base_path) if patient_info['patient'] in seg_info][0]
dicom_imgs = sitk_read_dcmseries(ori_path)
segs = sitk.ReadImage(seg_path)
segs = sitk.GetArrayFromImage(segs)
if len(segs)!=len(dicom_imgs):
print(index, 'img and seg not match')
for slice in range(len(segs)):
if segs[slice].any()==True:
img = dicom_imgs[slice]
seg = segs[slice]
print(slice, index)
if seg.shape != (512, 512):
print(index, 'shape not normal', seg.shape)
#####shift 调整至【0，2048】
window=[-1024, 1024]
img[np.where(img < window[0])] = window[0]
img[np.where(img > window[1])] = window[1]
img -= window[0]
print('image_max is ',img.max(),'min is', img.min())
hosp_patient_name = patient_info['name'] + '.npy'
np.save(os.path.join(newimg_path, hosp_patient_name), img)
np.save(os.path.join(newseg_path, hosp_patient_name), seg)
plt.imsave(os.path.join(newpngs_path, patient_info['name'] +'.png'), img, cmap = 'gray')
################## dpyy
base_path = 'data/data_original/dpyy'
seg_base_path = 'data/data_original/dpyy_seg'
patients_info = exter2_data[exter2_data['hospital']=='dpyy']
for index, patient_info in patients_info.iterrows():
####大坪医院多个'V'文件夹
ori_path = os.path.join(base_path, patient_info['patient'], 'V')
seg_path = [os.path.join(seg_base_path, seg_info) for seg_info in os.listdir(seg_base_path) if patient_info['patient'] in seg_info][0]
dicom_imgs = sitk_read_dcmseries(ori_path)
segs = sitk.ReadImage(seg_path)
segs = sitk.GetArrayFromImage(segs)
if len(segs)!=len(dicom_imgs):
print(index, 'img and seg not match')
for slice in range(len(segs)):
if segs[slice].any()==True:
img = dicom_imgs[slice]
seg = segs[slice]
print(slice, index)
if seg.shape != (512, 512):
print(index, 'shape not normal', seg.shape)
img = img[-512:,:]
seg = seg[-512:,:]
#####shift 调整至【0，2048】
window=[-1024, 1024]
img[np.where(img < window[0])] = window[0]
img[np.where(img > window[1])] = window[1]
img -= window[0]
print('image_max is ',img.max(),'min is', img.min())
hosp_patient_name = patient_info['name'] + '.npy'
np.save(os.path.join(newimg_path, hosp_patient_name), img)
np.save(os.path.join(newseg_path, hosp_patient_name), seg)
plt.imsave(os.path.join(newpngs_path, patient_info['name'] +'.png'), img, cmap = 'gray')