-
Notifications
You must be signed in to change notification settings - Fork 3
/
3D-CRF.py
164 lines (121 loc) · 5.7 KB
/
3D-CRF.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
"""
三维全连接条件随机场后处理优化
"""
import os
import sys
sys.path.append(os.path.split(sys.path[0])[0])
import collections
import numpy as np
import pandas as pd
from tqdm import tqdm
import SimpleITK as sitk
import pydensecrf.densecrf as dcrf
from pydensecrf.utils import create_pairwise_bilateral, create_pairwise_gaussian, unary_from_softmax
from utils.metrics import Metirc
import parameter as para
file_name = [] # 文件名称
# 定义评价指标
liver_score = collections.OrderedDict()
liver_score['dice'] = []
liver_score['jacard'] = []
liver_score['voe'] = []
liver_score['rvd'] = []
liver_score['fnr'] = []
liver_score['fpr'] = []
liver_score['assd'] = []
liver_score['rmsd'] = []
liver_score['msd'] = []
# 为了计算dice_global定义的两个变量
dice_intersection = 0.0
dice_union = 0.0
for file_index, file in enumerate(os.listdir(para.test_liver_ct_path)):
print('file index:', file_index, file, '--------------------------------------')
file_name.append(file)
ct = sitk.ReadImage(os.path.join(para.test_liver_ct_path, file), sitk.sitkInt16)
ct_array = sitk.GetArrayFromImage(ct)
pred = sitk.ReadImage(os.path.join(para.pred_liver_path, file.replace('volume', 'pred')), sitk.sitkUInt8)
pred_array = sitk.GetArrayFromImage(pred)
seg = sitk.ReadImage(os.path.join(para.test_liver_seg_path, file.replace('volume', 'segmentation')), sitk.sitkUInt8)
seg_array = sitk.GetArrayFromImage(seg)
seg_array[seg_array > 0] = 1
# 灰度截断
ct_array[ct_array > para.upper] = para.upper
ct_array[ct_array < para.lower] = para.lower
# 切割出预测结果部分，减少crf处理难度
z = np.any(pred_array, axis=(1, 2))
start_z, end_z = np.where(z)[0][[0, -1]]
y = np.any(pred_array, axis=(0, 1))
start_y, end_y = np.where(y)[0][[0, -1]]
x = np.any(pred_array, axis=(0, 2))
start_x, end_x = np.where(x)[0][[0, -1]]
# 扩张
start_z = max(0, start_z - para.z_expand)
start_x = max(0, start_x - para.x_expand)
start_y = max(0, start_y - para.y_expand)
end_z = min(ct_array.shape[0], end_z + para.z_expand)
end_x = min(ct_array.shape[1], end_x + para.x_expand)
end_y = min(ct_array.shape[2], end_y + para.y_expand)
new_ct_array = ct_array[start_z: end_z, start_x: end_x, start_y: end_y]
new_pred_array = pred_array[start_z: end_z, start_x: end_x, start_y: end_y]
print('old shape', ct_array.shape)
print('new shape', new_ct_array.shape)
print('shrink to:', np.prod(new_ct_array.shape) / np.prod(ct_array.shape), '%')
# 定义条件随机场
n_labels = 2
d = dcrf.DenseCRF(np.prod(new_ct_array.shape), n_labels)
# 获取一元势
unary = np.zeros_like(new_pred_array, dtype=np.float32)
unary[new_pred_array == 0] = 0.1
unary[new_pred_array == 1] = 0.9
U = np.stack((1 - unary, unary), axis=0)
d.setUnaryEnergy(unary_from_softmax(U))
# 获取二元势
# This creates the color-independent features and then add them to the CRF
feats = create_pairwise_gaussian(sdims=(para.s1, para.s1, para.s1), shape=new_ct_array.shape)
d.addPairwiseEnergy(feats, compat=3, kernel=dcrf.DIAG_KERNEL, normalization=dcrf.NORMALIZE_SYMMETRIC)
# This creates the color-dependent features and then add them to the CRF
feats = create_pairwise_bilateral(sdims=(para.s2, para.s2, para.s2), schan=(para.s3,), img=new_ct_array)
d.addPairwiseEnergy(feats, compat=10, kernel=dcrf.DIAG_KERNEL, normalization=dcrf.NORMALIZE_SYMMETRIC)
# 进行推理
Q, tmp1, tmp2 = d.startInference()
for i in tqdm(range(para.max_iter)):
# print("KL-divergence at {}: {}".format(i, d.klDivergence(Q)))
d.stepInference(Q, tmp1, tmp2)
# 获取预测标签结果
MAP = np.argmax(np.array(Q), axis=0).reshape(new_pred_array.shape)
liver_seg = np.zeros_like(seg_array, dtype=np.uint8)
liver_seg[start_z: end_z, start_x: end_x, start_y: end_y] = MAP.astype(np.uint8)
# 计算分割评价指标
liver_metric = Metirc(seg_array, liver_seg, ct.GetSpacing())
liver_score['dice'].append(liver_metric.get_dice_coefficient()[0])
liver_score['jacard'].append(liver_metric.get_jaccard_index())
liver_score['voe'].append(liver_metric.get_VOE())
liver_score['rvd'].append(liver_metric.get_RVD())
liver_score['fnr'].append(liver_metric.get_FNR())
liver_score['fpr'].append(liver_metric.get_FPR())
liver_score['assd'].append(liver_metric.get_ASSD())
liver_score['rmsd'].append(liver_metric.get_RMSD())
liver_score['msd'].append(liver_metric.get_MSD())
dice_intersection += liver_metric.get_dice_coefficient()[1]
dice_union += liver_metric.get_dice_coefficient()[2]
# 将CRF后处理的结果保存为nii数据
pred_seg = sitk.GetImageFromArray(liver_seg)
pred_seg.SetDirection(ct.GetDirection())
pred_seg.SetOrigin(ct.GetOrigin())
pred_seg.SetSpacing(ct.GetSpacing())
sitk.WriteImage(pred_seg, os.path.join(para.crf_liver_path, file.replace('volume', 'crf')))
print('dice:', liver_score['dice'][-1])
print('--------------------------------------------------------------')
# 将评价指标写入到exel中
liver_data = pd.DataFrame(liver_score, index=file_name)
liver_statistics = pd.DataFrame(index=['mean', 'std', 'min', 'max'], columns=list(liver_data.columns))
liver_statistics.loc['mean'] = liver_data.mean()
liver_statistics.loc['std'] = liver_data.std()
liver_statistics.loc['min'] = liver_data.min()
liver_statistics.loc['max'] = liver_data.max()
writer = pd.ExcelWriter('./result_pred_xlsx/result-post-processing_five_five.xlsx')
liver_data.to_excel(writer, 'liver')
liver_statistics.to_excel(writer, 'liver_statistics')
writer.save()
# 打印dice global
print('dice global:', dice_intersection / dice_union)