-
Notifications
You must be signed in to change notification settings - Fork 0
/
coarse_fusion.py
249 lines (230 loc) · 10.5 KB
/
coarse_fusion.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
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
import numpy as np
import os
import sys
import time
from utils import *
data_path = sys.argv[1]
current_fold = int(sys.argv[2])
organ_number = int(sys.argv[3])
low_range = int(sys.argv[4])
high_range = int(sys.argv[5])
slice_threshold = float(sys.argv[6])
slice_thickness = int(sys.argv[7])
organ_ID = int(sys.argv[8])
GPU_ID = int(sys.argv[9])
learning_rate1 = float(sys.argv[10])
learning_rate_m1 = int(sys.argv[11])
learning_rate2 = float(sys.argv[12])
learning_rate_m2 = int(sys.argv[13])
crop_margin = int(sys.argv[14])
result_path = os.path.join(result_path, 'coarse_testing_' + \
sys.argv[10] + 'x' + str(learning_rate_m1) + ',' + str(crop_margin))
epoch = 'e' + sys.argv[15] + sys.argv[16] + sys.argv[17] + sys.argv[18]
epoch_list = [epoch]
threshold = float(sys.argv[19])
timestamp = {}
timestamp['X'] = sys.argv[20]
timestamp['Y'] = sys.argv[21]
timestamp['Z'] = sys.argv[22]
volume_list = open(testing_set_filename(current_fold), 'r').read().splitlines()
while volume_list[len(volume_list) - 1] == '':
volume_list.pop()
result_name_ = {}
result_directory_ = {}
for plane in ['X', 'Y', 'Z']:
result_name__ = result_name_from_timestamp(result_path, current_fold, \
plane, 'I', slice_thickness, organ_ID, volume_list, timestamp[plane])
if result_name__ == '':
exit(' Error: no valid result directories are detected!')
result_directory__ = os.path.join(result_path, result_name__, 'volumes')
print(' Result directory for plane ' + plane + ': ' + result_directory__ + ' .')
if result_name__.startswith('FD'):
index_ = result_name__.find(':')
result_name__ = result_name__[index_ + 1: ]
result_name_[plane] = result_name__
result_directory_[plane] = result_directory__
DSC_X = np.zeros((len(volume_list)))
DSC_Y = np.zeros((len(volume_list)))
DSC_Z = np.zeros((len(volume_list)))
DSC_F1 = np.zeros((len(volume_list)))
DSC_F2 = np.zeros((len(volume_list)))
DSC_F3 = np.zeros((len(volume_list)))
DSC_F1P = np.zeros((len(volume_list)))
DSC_F2P = np.zeros((len(volume_list)))
DSC_F3P = np.zeros((len(volume_list)))
result_name = 'FD' + str(current_fold) + ':' + 'fusion:' + result_name_['X'] + ',' + \
result_name_['Y'] + ',' + result_name_['Z'] + '_' + epoch + ',' + str(threshold)
result_directory = os.path.join(result_path, result_name, 'volumes')
if not os.path.exists(result_directory):
os.makedirs(result_directory)
result_file = os.path.join(result_path, result_name, 'results.txt')
output = open(result_file, 'w')
output.close()
output = open(result_file, 'a+')
output.write('Fusing results of ' + str(len(epoch_list)) + ' snapshots:\n')
output.close()
for i in range(len(volume_list)):
start_time = time.time()
print('Testing ' + str(i + 1) + ' out of ' + str(len(volume_list)) + ' testcases.')
output = open(result_file, 'a+')
output.write(' Testcase ' + str(i + 1) + ':\n')
output.close()
s = volume_list[i].split(' ')
label = np.load(s[2])
label = is_organ(label, organ_ID).astype(np.uint8)
for plane in ['X', 'Y', 'Z']:
volume_file = volume_filename_fusion(result_directory, plane, i)
pred = np.zeros(label.shape, dtype = np.float32)
for t in range(len(epoch_list)):
volume_file_ = volume_filename_testing(result_directory_[plane], epoch_list[t], i)
pred += np.load(volume_file_)['volume']
pred_ = (pred >= threshold * 255 * len(epoch_list))
if not os.path.isfile(volume_file):
np.savez_compressed(volume_file, volume = pred_)
DSC_, inter_sum, pred_sum, label_sum = DSC_computation(label, pred_)
print(' DSC_' + plane + ' = 2 * ' + str(inter_sum) + ' / (' + \
str(pred_sum) + ' + ' + str(label_sum) + ') = ' + str(DSC_) + ' .')
output = open(result_file, 'a+')
output.write(' DSC_' + plane + ' = 2 * ' + str(inter_sum) + ' / (' + \
str(pred_sum) + ' + ' + str(label_sum) + ') = ' + str(DSC_) + ' .\n')
output.close()
if pred_sum == 0 and label_sum == 0:
DSC_ = 0
pred /= (255 * len(epoch_list))
if plane == 'X':
pred_X = pred
DSC_X[i] = DSC_
elif plane == 'Y':
pred_Y = pred
DSC_Y[i] = DSC_
elif plane == 'Z':
pred_Z = pred
DSC_Z[i] = DSC_
volume_file_F1 = volume_filename_fusion(result_directory, 'F1', i)
volume_file_F2 = volume_filename_fusion(result_directory, 'F2', i)
volume_file_F3 = volume_filename_fusion(result_directory, 'F3', i)
if not os.path.isfile(volume_file_F1) or not os.path.isfile(volume_file_F2) or \
not os.path.isfile(volume_file_F3):
pred_total = pred_X + pred_Y + pred_Z
if os.path.isfile(volume_file_F1):
pred_F1 = np.load(volume_file_F1)['volume'].astype(np.uint8)
else:
pred_F1 = (pred_total >= 0.5).astype(np.uint8)
np.savez_compressed(volume_file_F1, volume = pred_F1)
DSC_F1[i], inter_sum, pred_sum, label_sum = DSC_computation(label, pred_F1)
print(' DSC_F1 = 2 * ' + str(inter_sum) + ' / (' + str(pred_sum) + ' + ' \
+ str(label_sum) + ') = ' + str(DSC_F1[i]) + ' .')
output = open(result_file, 'a+')
output.write(' DSC_F1 = 2 * ' + str(inter_sum) + ' / (' + \
str(pred_sum) + ' + ' + str(label_sum) + ') = ' + str(DSC_F1[i]) + ' .\n')
output.close()
if pred_sum == 0 and label_sum == 0:
DSC_F1[i] = 0
if os.path.isfile(volume_file_F2):
pred_F2 = np.load(volume_file_F2)['volume'].astype(np.uint8)
else:
pred_F2 = (pred_total >= 1.5).astype(np.uint8)
np.savez_compressed(volume_file_F2, volume = pred_F2)
DSC_F2[i], inter_sum, pred_sum, label_sum = DSC_computation(label, pred_F2)
print(' DSC_F2 = 2 * ' + str(inter_sum) + ' / (' + str(pred_sum) + ' + ' + \
str(label_sum) + ') = ' + str(DSC_F2[i]) + ' .')
output = open(result_file, 'a+')
output.write(' DSC_F2 = 2 * ' + str(inter_sum) + ' / (' + \
str(pred_sum) + ' + ' + str(label_sum) + ') = ' + str(DSC_F2[i]) + ' .\n')
output.close()
if pred_sum == 0 and label_sum == 0:
DSC_F2[i] = 0
if os.path.isfile(volume_file_F3):
pred_F3 = np.load(volume_file_F3)['volume'].astype(np.uint8)
else:
pred_F3 = (pred_total >= 2.5).astype(np.uint8)
np.savez_compressed(volume_file_F3, volume = pred_F3)
DSC_F3[i], inter_sum, pred_sum, label_sum = DSC_computation(label, pred_F3)
print(' DSC_F3 = 2 * ' + str(inter_sum) + ' / (' + str(pred_sum) + ' + ' + \
str(label_sum) + ') = ' + str(DSC_F3[i]) + ' .')
output = open(result_file, 'a+')
output.write(' DSC_F3 = 2 * ' + str(inter_sum) + ' / (' + \
str(pred_sum) + ' + ' + str(label_sum) + ') = ' + str(DSC_F3[i]) + ' .\n')
output.close()
if pred_sum == 0 and label_sum == 0:
DSC_F3[i] = 0
volume_file_F1P = volume_filename_fusion(result_directory, 'F1P', i)
volume_file_F2P = volume_filename_fusion(result_directory, 'F2P', i)
volume_file_F3P = volume_filename_fusion(result_directory, 'F3P', i)
S = pred_F3
if (S.sum() == 0):
S = pred_F2
if (S.sum() == 0):
S = pred_F1
if os.path.isfile(volume_file_F1P):
pred_F1P = np.load(volume_file_F1P)['volume'].astype(np.uint8)
else:
pred_F1P = post_processing(pred_F1, S, 0.5, organ_ID)
np.savez_compressed(volume_file_F1P, volume = pred_F1P)
DSC_F1P[i], inter_sum, pred_sum, label_sum = DSC_computation(label, pred_F1P)
print(' DSC_F1P = 2 * ' + str(inter_sum) + ' / (' + str(pred_sum) + ' + ' + \
str(label_sum) + ') = ' + str(DSC_F1P[i]) + ' .')
output = open(result_file, 'a+')
output.write(' DSC_F1P = 2 * ' + str(inter_sum) + ' / (' + \
str(pred_sum) + ' + ' + str(label_sum) + ') = ' + str(DSC_F1P[i]) + ' .\n')
output.close()
if pred_sum == 0 and label_sum == 0:
DSC_F1P[i] = 0
if os.path.isfile(volume_file_F2P):
pred_F2P = np.load(volume_file_F2P)['volume'].astype(np.uint8)
else:
pred_F2P = post_processing(pred_F2, S, 0.5, organ_ID)
np.savez_compressed(volume_file_F2P, volume = pred_F2P)
DSC_F2P[i], inter_sum, pred_sum, label_sum = DSC_computation(label, pred_F2P)
print(' DSC_F2P = 2 * ' + str(inter_sum) + ' / (' + str(pred_sum) + ' + ' + \
str(label_sum) + ') = ' + str(DSC_F2P[i]) + ' .')
output = open(result_file, 'a+')
output.write(' DSC_F2P = 2 * ' + str(inter_sum) + ' / (' + \
str(pred_sum) + ' + ' + str(label_sum) + ') = ' + str(DSC_F2P[i]) + ' .\n')
output.close()
if pred_sum == 0 and label_sum == 0:
DSC_F2P[i] = 0
if os.path.isfile(volume_file_F3P):
pred_F3P = np.load(volume_file_F3P)['volume'].astype(np.uint8)
else:
pred_F3P = post_processing(pred_F3, S, 0.5, organ_ID)
np.savez_compressed(volume_file_F3P, volume = pred_F3P)
DSC_F3P[i], inter_sum, pred_sum, label_sum = DSC_computation(label, pred_F3P)
print(' DSC_F3P = 2 * ' + str(inter_sum) + ' / (' + str(pred_sum) + ' + ' + \
str(label_sum) + ') = ' + str(DSC_F3P[i]) + ' .')
output = open(result_file, 'a+')
output.write(' DSC_F3P = 2 * ' + str(inter_sum) + ' / (' + \
str(pred_sum) + ' + ' + str(label_sum) + ') = ' + str(DSC_F3P[i]) + ' .\n')
output.close()
if pred_sum == 0 and label_sum == 0:
DSC_F3P[i] = 0
pred_X = None
pred_Y = None
pred_Z = None
pred_F1 = None
pred_F2 = None
pred_F3 = None
pred_F1P = None
pred_F2P = None
pred_F3P = None
output = open(result_file, 'a+')
print('Average DSC_X = ' + str(np.mean(DSC_X)) + ' .')
output.write('Average DSC_X = ' + str(np.mean(DSC_X)) + ' .\n')
print('Average DSC_Y = ' + str(np.mean(DSC_Y)) + ' .')
output.write('Average DSC_Y = ' + str(np.mean(DSC_Y)) + ' .\n')
print('Average DSC_Z = ' + str(np.mean(DSC_Z)) + ' .')
output.write('Average DSC_Z = ' + str(np.mean(DSC_Z)) + ' .\n')
print('Average DSC_F1 = ' + str(np.mean(DSC_F1)) + ' .')
output.write('Average DSC_F1 = ' + str(np.mean(DSC_F1)) + ' .\n')
print('Average DSC_F2 = ' + str(np.mean(DSC_F2)) + ' .')
output.write('Average DSC_F2 = ' + str(np.mean(DSC_F2)) + ' .\n')
print('Average DSC_F3 = ' + str(np.mean(DSC_F3)) + ' .')
output.write('Average DSC_F3 = ' + str(np.mean(DSC_F3)) + ' .\n')
print('Average DSC_F1P = ' + str(np.mean(DSC_F1P)) + ' .')
output.write('Average DSC_F1P = ' + str(np.mean(DSC_F1P)) + ' .\n')
print('Average DSC_F2P = ' + str(np.mean(DSC_F2P)) + ' .')
output.write('Average DSC_F2P = ' + str(np.mean(DSC_F2P)) + ' .\n')
print('Average DSC_F3P = ' + str(np.mean(DSC_F3P)) + ' .')
output.write('Average DSC_F3P = ' + str(np.mean(DSC_F3P)) + ' .\n')
output.close()
print('The fusion process is finished.')