-
Notifications
You must be signed in to change notification settings - Fork 0
/
Copy pathsummary.py
144 lines (96 loc) · 4.02 KB
/
summary.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
import os
import numpy as np
import SimpleITK as sitk
from os.path import join
from utils.file_utils import save_json, load_json
import medpy.metric.binary as metric
'''
dir:
predict
---Unet
--raw
--postprocess
---AttUnet
--raw
--postprocess
...
---Other models
--raw
--postprocess
---raw_summary.json
---postprocess_summary.json
'''
def check_spacing(spacing_gt, spacing_pre):
same_spac = np.all(np.isclose(spacing_gt, spacing_pre))
if not same_spac:
print("the spacing does not match between the images")
print(spacing_gt)
print(spacing_pre)
def get_metric_dic(pre_path, gt_path):
res_dic = {}
for case in os.listdir(gt_path):
case_name = case.split('_label')[0]
gt = sitk.ReadImage(join(gt_path, case))
pre = sitk.ReadImage(join(pre_path, case.replace("_label", "")))
spacing = np.array(gt.GetSpacing())[[2, 1, 0]]
# check_spacing(spacing_gt, spacing_pre)
gt = sitk.GetArrayFromImage(gt)
pre = sitk.GetArrayFromImage(pre)
metric_dic = {}
metric_dic['dice'] = metric.dc(gt, pre)
metric_dic['jaccard'] = metric.jc(gt, pre)
metric_dic['precision'] = metric.precision(gt, pre)
metric_dic['recall'] = metric.recall(gt, pre)
metric_dic['specificity'] = metric.specificity(gt, pre)
metric_dic['hd'] = metric.hd(gt, pre, voxelspacing=spacing)
metric_dic['hd95'] = metric.hd95(gt, pre, voxelspacing=spacing)
metric_dic['assd'] = metric.assd(gt, pre, voxelspacing=spacing)
metric_dic['asd'] = metric.asd(gt, pre, voxelspacing=spacing)
# metric_dic['sensitivity'] = metric.sensitivity(gt, pre) # = recall
# metric_dic['true_positive_rate'] = metric.true_positive_rate(gt, pre) # = recall
# metric_dic['ravd'] = metric.ravd(gt, pre)
# metric_dic['volume_correlation'] = metric.volume_correlation(gt, pre)
res_dic[case_name] = metric_dic
print('Case: {} calculated done...'.format(case_name))
avg = {}
for name, case_metric in res_dic.items():
for key, value in case_metric.items():
if key not in avg.keys():
avg[key] = value
else:
avg[key] += value
for key, value in avg.items():
avg[key] = avg[key] / len(res_dic.keys())
res_dic['Average'] = avg
return res_dic
def get_summary(predict_path, gt_path, model_name):
save_raw_dir = join(predict_path, 'raw_summary.json')
save_postprocess_dir = join(predict_path, 'postprocess_summary.json')
if not os.path.exists(save_raw_dir) or not os.path.exists(save_postprocess_dir):
raw_metric_dic = {}
postprocess_metric_dic = {}
else:
raw_metric_dic = load_json(save_raw_dir)
postprocess_metric_dic = load_json(save_postprocess_dir)
for model in os.listdir(predict_path):
if model not in model_name:
continue
model_path = join(predict_path, model)
raw_path = join(model_path, 'raw')
postprocess_path = join(model_path, 'postprocess')
# if not os.path.exists(postprocess_path):
# os.mkdir(postprocess_path)
# print("postprocess dir maked successfully...")
print('Begin {} raw calculation...'.format(model))
raw_metric_dic[model] = get_metric_dic(raw_path, gt_path)
print('Begin {} postprocess calculation...'.format(model))
postprocess_metric_dic[model] = get_metric_dic(postprocess_path, gt_path)
print('Model: {} calculated done...'.format(model))
save_json(raw_metric_dic, save_raw_dir)
save_json(postprocess_metric_dic, save_postprocess_dir)
print('Summary json file has been saved done...')
if __name__ == '__main__':
predict_path = '/data1/zfx/data/BileDuct/predict'
gt_path = '/data1/zfx/data/BileDuct/raw_data/labelsTs'
model_name = ['TAGNet']
get_summary(predict_path, gt_path, model_name)