|
1
|
|
|
""" |
|
2
|
|
|
Download the demo data |
|
3
|
|
|
""" |
|
4
|
|
|
import os |
|
5
|
|
|
import shutil |
|
6
|
|
|
import zipfile |
|
7
|
|
|
|
|
8
|
|
|
from tensorflow.keras.utils import get_file |
|
9
|
|
|
|
|
10
|
|
|
PROJECT_DIR = "demos/paired_mrus_prostate" |
|
11
|
|
|
os.chdir(PROJECT_DIR) |
|
12
|
|
|
|
|
13
|
|
|
DATA_PATH = "dataset" |
|
14
|
|
|
ZIP_PATH = "example-data-mrus" |
|
15
|
|
|
ORIGIN = "https://github.com/yipenghu/example-data/archive/mrus.zip" |
|
16
|
|
|
|
|
17
|
|
|
zip_file = ZIP_PATH + ".zip" |
|
18
|
|
|
get_file(os.path.abspath(zip_file), ORIGIN) |
|
19
|
|
|
with zipfile.ZipFile(zip_file, "r") as zf: |
|
20
|
|
|
zf.extractall() |
|
21
|
|
|
|
|
22
|
|
|
if os.path.exists(DATA_PATH): |
|
23
|
|
|
shutil.rmtree(DATA_PATH) |
|
24
|
|
|
os.rename(ZIP_PATH, DATA_PATH) |
|
25
|
|
|
os.remove(zip_file) |
|
26
|
|
|
|
|
27
|
|
|
print("\nMR and ultrasound data downloaded: %s." % os.path.abspath(DATA_PATH)) |
|
28
|
|
|
|
|
29
|
|
|
# now split the data in to num_part partitions |
|
30
|
|
|
num_part = 11 |
|
31
|
|
|
|
|
32
|
|
|
data_types = ["moving_images", "moving_labels", "fixed_images", "fixed_labels"] |
|
33
|
|
|
filenames = [sorted(os.listdir(os.path.join(DATA_PATH, fn))) for fn in data_types] |
|
34
|
|
|
num_files = [len(x) for x in filenames] |
|
35
|
|
|
if len(set(num_files)) != 1: |
|
36
|
|
|
raise ValueError( |
|
37
|
|
|
"Number of data are not the same between moving/fixed/images/labels. " |
|
38
|
|
|
"Please run this download script again." |
|
39
|
|
|
) |
|
40
|
|
|
num_data = num_files[0] |
|
41
|
|
|
|
|
42
|
|
|
for idx in range(num_part): # create partition folders |
|
43
|
|
|
os.makedirs(os.path.join(DATA_PATH, "part%02d" % idx)) |
|
44
|
|
|
for fn in data_types: |
|
45
|
|
|
os.makedirs(os.path.join(DATA_PATH, "part%02d" % idx, fn)) |
|
46
|
|
|
|
|
47
|
|
|
for idx in range(num_data): # copy all files to part folders |
|
48
|
|
|
for ifn in range(len(data_types)): |
|
49
|
|
|
os.rename( |
|
50
|
|
|
os.path.join(DATA_PATH, data_types[ifn], filenames[ifn][idx]), |
|
51
|
|
|
os.path.join( |
|
52
|
|
|
DATA_PATH, |
|
53
|
|
|
"part%02d" % (idx % num_part), |
|
54
|
|
|
data_types[ifn], |
|
55
|
|
|
filenames[ifn][idx], |
|
56
|
|
|
), |
|
57
|
|
|
) |
|
58
|
|
|
|
|
59
|
|
|
for fn in data_types: # remove the old type folders |
|
60
|
|
|
shutil.rmtree(os.path.join(DATA_PATH, fn)) |
|
61
|
|
|
|
|
62
|
|
|
print("All data are partitioned into %d folders." % num_part) |
|
63
|
|
|
|
|
64
|
|
|
## now download the pre-trained model |
|
65
|
|
|
MODEL_PATH = os.path.join(DATA_PATH, "pretrained") |
|
66
|
|
|
if os.path.exists(MODEL_PATH): |
|
67
|
|
|
shutil.rmtree(MODEL_PATH) |
|
68
|
|
|
os.mkdir(MODEL_PATH) |
|
69
|
|
|
|
|
70
|
|
|
ZIP_PATH = "checkpoint" |
|
71
|
|
|
ORIGIN = "https://github.com/DeepRegNet/deepreg-model-zoo/raw/master/demo/paired_mrus_prostate/20210110.zip" |
|
72
|
|
|
|
|
73
|
|
|
zip_file = os.path.join(MODEL_PATH, ZIP_PATH + ".zip") |
|
74
|
|
|
get_file(os.path.abspath(zip_file), ORIGIN) |
|
75
|
|
|
with zipfile.ZipFile(zip_file, "r") as zf: |
|
76
|
|
|
zf.extractall(path=MODEL_PATH) |
|
77
|
|
|
os.remove(zip_file) |
|
78
|
|
|
|
|
79
|
|
|
print( |
|
80
|
|
|
"Pre-trained model is downloaded and unzipped in %s." % os.path.abspath(MODEL_PATH) |
|
81
|
|
|
) |
|
82
|
|
|
|
DeepRegNet /
DeepReg
