forked from Taurusmumu/Sonnet_pytorch
-
Notifications
You must be signed in to change notification settings - Fork 0
/
convert_format.py
executable file
·102 lines (84 loc) · 3.3 KB
/
convert_format.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
"""convert_format.py.
Used to convert output to a format that can be used for visualisaton with QuPath.
Note, this is only used for tile segmentation results; not WSI.
"""
import os
import re
import glob
import json
import pathlib
import numpy as np
import shutil
from misc.utils import rm_n_mkdir, mkdir
####
def to_qupath(file_path, nuc_pos_list, nuc_type_list, type_info_dict):
"""
For QuPath v0.2.3
"""
def rgb2int(rgb):
r, g, b = rgb
return (r << 16) + (g << 8) + b
nuc_pos_list = np.array(nuc_pos_list)
nuc_type_list = np.array(nuc_type_list)
assert nuc_pos_list.shape[0] == nuc_type_list.shape[0]
with open(file_path, "w") as fptr:
fptr.write("x\ty\tclass\tname\tcolor\n")
nr_nuc = nuc_pos_list.shape[0]
for idx in range(nr_nuc):
nuc_type = nuc_type_list[idx]
nuc_pos = nuc_pos_list[idx]
type_name = type_info_dict[nuc_type][0]
type_color = type_info_dict[nuc_type][1]
type_color = rgb2int(type_color) # color in qupath format
fptr.write(
"{x}\t{y}\t{type_class}\t{type_name}\t{type_color}\n".format(
x=nuc_pos[0],
y=nuc_pos[1],
type_class="",
type_name=type_name,
type_color=type_color,
)
)
return
####
if __name__ == "__main__":
target_format = "qupath"
# to rescale the coordinate set to match with lv0 mag of the wsi
scale_factor = 1.0
root_dir = "dataset/dummy/out/"
# to define the name, and color conversion code for each target format
type_info_dict = {
0: ("nolabe", (0, 0, 0)), # no label
1: ("neopla", (255, 0, 0)), # neoplastic
2: ("inflam", (0, 255, 0)), # inflamm
3: ("connec", (0, 0, 255)), # connective
4: ("necros", (255, 255, 0)), # dead
5: ("no-neo", (255, 165, 0)), # non-neoplastic epithelial
}
patterning = lambda x: re.sub("([\[\]])", "[\\1]", x)
code_name_list = glob.glob(patterning("%s/*.json" % root_dir))
code_name_list = [pathlib.Path(v).stem for v in code_name_list]
code_name_list.sort()
output_dir = root_dir
for code_name in code_name_list[:]:
nuc_info_path = "%s/%s.json" % (root_dir, code_name)
if not os.path.exists(nuc_info_path):
continue
print(code_name)
with open(nuc_info_path, "r") as handle:
info_dict = json.load(handle)["nuc"]
# from json to array
new_info_dict = {}
for inst_id, inst_info in info_dict.items():
new_inst_info = {}
for info_name, info_value in inst_info.items():
if isinstance(info_value, list):
info_value = np.array(info_value)
info_value = info_value * scale_factor
info_value = info_value.astype(np.int32)
new_inst_info[info_name] = info_value
new_info_dict[inst_id] = new_inst_info
centroid_list = np.array([v["centroid"] for v in list(new_info_dict.values())])
type_list = np.array([v["type"] for v in list(new_info_dict.values())])
save_path = "%s/%s.tsv" % (output_dir, code_name)
to_qupath(save_path, centroid_list, type_list, type_info_dict)