-
Notifications
You must be signed in to change notification settings - Fork 10
/
Copy pathutil_script_prostateMRI.py
196 lines (147 loc) · 7.19 KB
/
util_script_prostateMRI.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 shutil
import numpy as np
import matplotlib.pyplot as plt
import os
import nibabel as nib
from torch.nn import functional as F
import pickle
import random
import pandas as pd
from tqdm import tqdm
import cv2
import SimpleITK as sitk
from scipy.ndimage.interpolation import zoom
import json
base_dir = "/mnt/weka/wekafs/rad-megtron/cchen"
def organize_data():
save_pth = base_dir + '/' + 'Dataset911_prostateD'
os.makedirs(save_pth+'/imagesTr', exist_ok=True)
os.makedirs(save_pth+'/labelsTr', exist_ok=True)
data_pth = base_dir + '/' + 'prostateD'
data_fd_list = os.listdir(data_pth)
data_fd_list = [data_fd for data_fd in data_fd_list if data_fd.endswith('.nii.gz') and 'segmentation' not in data_fd]
data_fd_list.sort()
for data_fd in data_fd_list:
patient_ID = '00' + data_fd[4:6]
mask_obj = nib.load(base_dir + '/' + 'prostateD/'+data_fd.replace('.nii.gz', '_segmentation.nii.gz'))
mask_arr = mask_obj.get_fdata()
mask_arr[mask_arr>1] = 1
new_mask_obj = nib.Nifti1Image(mask_arr, mask_obj.affine, header=mask_obj.header)
nib.save(new_mask_obj, save_pth+'/labelsTr/prostateD_'+patient_ID+'.nii.gz')
shutil.copy(base_dir + '/' + 'prostateD/'+data_fd, save_pth+'/imagesTr/prostateD_'+patient_ID+'_0000.nii.gz')
def get_3D_2D_all_5slice():
save_pth = base_dir + '/prostateD/2D_all_5slice'
data_pth_all = [base_dir + '/Dataset911_prostateD',
]
for data_pth in data_pth_all:
data_fd_list = os.listdir(data_pth+'/imagesTr')
data_fd_list = [data_fd for data_fd in data_fd_list if data_fd.endswith('.nii.gz')]
data_fd_list.sort()
cnt = 0
for data_fd_indx, data_fd in enumerate(data_fd_list):
case_id = data_fd.split('_')[1]
if not os.path.exists(save_pth+'/'+case_id):
os.makedirs(save_pth+'/'+case_id)
os.mkdir(save_pth+'/'+case_id+'/images')
os.mkdir(save_pth+'/'+case_id+'/masks')
img_obj = nib.load(data_pth + '/imagesTr/' + data_fd)
img_arr = img_obj.get_fdata()
mask_obj = nib.load(data_pth + '/labelsTr/' + data_fd.replace('_0000.nii.gz', '.nii.gz'))
mask_arr = mask_obj.get_fdata()
img_arr = np.float32(img_arr)
mask_arr = np.float32(mask_arr)
high = np.quantile(img_arr, 0.99)
low = np.min(img_arr)
img_arr = np.where(img_arr > high, high, img_arr)
lungwin = np.array([low * 1., high * 1.])
img_arr = (img_arr - lungwin[0]) / (lungwin[1] - lungwin[0])
h, w = img_arr.shape[0], img_arr.shape[1]
out_h, out_w = 512, 512
if h != 512 or w !=512:
print(h, w)
img_arr = zoom(img_arr, (out_h / h, out_w / w, 1.0), order=3)
mask_arr = zoom(mask_arr, (out_h / h, out_w / w, 1.0), order=0)
img_arr = np.concatenate((img_arr[:, :, 0:1], img_arr[:, :, 0:1], img_arr, img_arr[:, :, -1:], img_arr[:, :, -1:]), axis=-1)
mask_arr = np.concatenate((mask_arr[:, :, 0:1], mask_arr[:, :, 0:1], mask_arr, mask_arr[:, :, -1:], mask_arr[:, :, -1:]), axis=-1)
for slice_indx in range(2, img_arr.shape[2]-2):
slice_arr = img_arr[:,:,slice_indx-2: slice_indx+3]
slice_arr = np.flip(np.rot90(slice_arr, k=1, axes=(0, 1)), axis=1)
mask_arr_2D = mask_arr[:,:,slice_indx-2: slice_indx+3]
mask_arr_2D = np.flip(np.rot90(mask_arr_2D, k=1, axes=(0, 1)), axis=1)
with open(save_pth+'/'+case_id+'/images'+'/2Dimage_'+'{:04d}'.format(slice_indx-2)+'.pkl', 'wb') as file:
pickle.dump(slice_arr, file)
with open(save_pth+'/'+case_id+'/masks'+'/2Dmask_'+'{:04d}'.format(slice_indx-2)+'.pkl', 'wb') as file:
pickle.dump(mask_arr_2D, file)
cnt += 1
def get_csv():
save_pth = base_dir + '/prostateD/2D_all_5slice'
training_csv = save_pth+'/training.csv'
validation_csv = save_pth+'/validation.csv'
test_csv = save_pth+'/test.csv'
all_csv = save_pth+'/all.csv'
data_fd_list = os.listdir(save_pth)
data_fd_list = [data_fd for data_fd in data_fd_list if data_fd.startswith('00') and '.' not in data_fd]
random.shuffle(data_fd_list)
random.shuffle(data_fd_list)
random.shuffle(data_fd_list)
random.shuffle(data_fd_list)
random.shuffle(data_fd_list)
test_fd_list = ['0001', '0032', '0034']
training_fd_list = list(set(data_fd_list)-set(test_fd_list))
validation_fd_list = random.sample(test_fd_list, min(len(test_fd_list), 4))
path_list_all = []
for data_fd in data_fd_list:
slice_list = os.listdir(save_pth+'/'+data_fd+'/images')
slice_pth_list = [data_fd+'/images/'+slice for slice in slice_list]
path_list_all = path_list_all + slice_pth_list
random.shuffle(path_list_all)
random.shuffle(path_list_all)
random.shuffle(path_list_all)
random.shuffle(path_list_all)
random.shuffle(path_list_all)
df = pd.DataFrame(path_list_all, columns=['image_pth'])
df['mask_pth'] = path_list_all
df['mask_pth'] = df['mask_pth'].apply(lambda x: x.replace('/images/2Dimage_', '/masks/2Dmask_'))
df.to_csv(all_csv, index=False)
path_list_all = []
for data_fd in training_fd_list:
slice_list = os.listdir(save_pth+'/'+data_fd+'/images')
slice_pth_list = [data_fd+'/images/'+slice for slice in slice_list]
path_list_all = path_list_all + slice_pth_list
random.shuffle(path_list_all)
random.shuffle(path_list_all)
random.shuffle(path_list_all)
random.shuffle(path_list_all)
random.shuffle(path_list_all)
df = pd.DataFrame(path_list_all, columns=['image_pth'])
df['mask_pth'] = path_list_all
df['mask_pth'] = df['mask_pth'].apply(lambda x: x.replace('/images/2Dimage_', '/masks/2Dmask_'))
df.to_csv(training_csv, index=False)
path_list_all = []
for data_fd in validation_fd_list:
slice_list = os.listdir(save_pth+'/'+data_fd+'/images')
slice_pth_list = [data_fd+'/images/'+slice for slice in slice_list]
path_list_all = path_list_all + slice_pth_list
random.shuffle(path_list_all)
random.shuffle(path_list_all)
random.shuffle(path_list_all)
random.shuffle(path_list_all)
random.shuffle(path_list_all)
df = pd.DataFrame(path_list_all, columns=['image_pth'])
df['mask_pth'] = path_list_all
df['mask_pth'] = df['mask_pth'].apply(lambda x: x.replace('/images/2Dimage_', '/masks/2Dmask_'))
df.to_csv(validation_csv, index=False)
path_list_all = []
for data_fd in test_fd_list:
slice_list = os.listdir(save_pth+'/'+data_fd+'/images')
slice_list.sort()
slice_pth_list = [data_fd+'/images/'+slice for slice in slice_list]
path_list_all = path_list_all + slice_pth_list
df = pd.DataFrame(path_list_all, columns=['image_pth'])
df['mask_pth'] = path_list_all
df['mask_pth'] = df['mask_pth'].apply(lambda x: x.replace('/images/2Dimage_', '/masks/2Dmask_'))
df.to_csv(test_csv, index=False)
if __name__=="__main__":
organize_data()
get_3D_2D_all_5slice()
get_csv()