-
Notifications
You must be signed in to change notification settings - Fork 2
/
Copy pathdata.py
119 lines (93 loc) · 4.78 KB
/
data.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
import os
import numpy as np
import torch
import torch.utils.data as udata
import SimpleITK as sitk
import pandas as pd
from augment import augment_data
def resample(image, spacing, size):
# Create the reference image
reference_origin = np.zeros(image.GetDimension())
reference_direction = np.identity(image.GetDimension()).flatten()
reference_image = sitk.Image(size, image.GetPixelIDValue())
reference_image.SetOrigin(reference_origin)
reference_image.SetSpacing(spacing)
reference_image.SetDirection(reference_direction)
# Transform which maps from the reference_image to the current image (output-to-input)
transform = sitk.AffineTransform(image.GetDimension())
transform.SetMatrix(image.GetDirection())
transform.SetTranslation(np.array(image.GetOrigin()) - reference_origin)
# Modify the transformation to align the centers of the original and reference image
reference_center = np.array(reference_image.TransformContinuousIndexToPhysicalPoint(np.array(reference_image.GetSize())/2.0))
centering_transform = sitk.TranslationTransform(image.GetDimension())
img_center = np.array(image.TransformContinuousIndexToPhysicalPoint(np.array(image.GetSize())/2.0))
centering_transform.SetOffset(np.array(transform.GetInverse().TransformPoint(img_center) - reference_center))
centered_transform = sitk.Transform(transform)
centered_transform.AddTransform(centering_transform)
# Using the linear interpolator
image_rs = sitk.Resample(image, reference_image, transform, sitk.sitkLinear, 0.0)
return image_rs
class Radpath(udata.Dataset):
def __init__(self, csv_file, data_path, shuffle):
self.image_path = data_path
self.df = pd.read_csv(csv_file)
if shuffle:
self.df = self.df.sample(frac=1).reset_index(drop=True)
def __len__(self):
return len(self.df)
def __getitem__(self, idx):
label_dict = {'G': 0, 'O': 1, 'A': 2}
label = label_dict[self.df.loc[idx, 'class']]
dataID = self.df.loc[idx, 'CPM_RadPath_2019_ID']
# read niigz file
T1 = sitk.ReadImage(os.path.join(self.image_path, dataID, dataID+'_t1.nii.gz'))
T1ce = sitk.ReadImage(os.path.join(self.image_path, dataID, dataID+'_t1ce.nii.gz'))
T2 = sitk.ReadImage(os.path.join(self.image_path, dataID, dataID+'_t2.nii.gz'))
FLAIR = sitk.ReadImage(os.path.join(self.image_path, dataID, dataID+'_flair.nii.gz'))
# resize
x_size, y_size, z_size = T1.GetSize()
input_size = [128, 128, 128]
spacing = [x_size / input_size[0], y_size / input_size[1], z_size / input_size[2]]
T1 = resample(T1, spacing=spacing, size=input_size)
T1ce = resample(T1ce, spacing=spacing, size=input_size)
T2 = resample(T2, spacing=spacing, size=input_size)
FLAIR = resample(FLAIR, spacing=spacing, size=input_size)
# convert to one batch of ndarray
T1 = sitk.GetArrayFromImage(T1).astype(np.float32)
T1ce = sitk.GetArrayFromImage(T1ce).astype(np.float32)
T2 = sitk.GetArrayFromImage(T2).astype(np.float32)
FLAIR = sitk.GetArrayFromImage(FLAIR).astype(np.float32)
image = np.stack((T1, T1ce, T2, FLAIR), 0)
# tensor
data = torch.from_numpy(image)
return data, label
class Radpath_test(udata.Dataset):
def __init__(self, csv_file, data_path):
self.image_path = data_path
self.df = pd.read_csv(csv_file)
def __len__(self):
return len(self.df)
def __getitem__(self, idx):
dataID = self.df.loc[idx, 'CPM_RadPath_2020_ID']
# read niigz file
T1 = sitk.ReadImage(os.path.join(self.image_path, dataID, dataID+'_t1.nii.gz'))
T1ce = sitk.ReadImage(os.path.join(self.image_path, dataID, dataID+'_t1ce.nii.gz'))
T2 = sitk.ReadImage(os.path.join(self.image_path, dataID, dataID+'_t2.nii.gz'))
FLAIR = sitk.ReadImage(os.path.join(self.image_path, dataID, dataID+'_flair.nii.gz'))
# resize
x_size, y_size, z_size = T1.GetSize()
input_size = [128, 128, 128]
spacing = [x_size / input_size[0], y_size / input_size[1], z_size / input_size[2]]
T1 = resample(T1, spacing=spacing, size=input_size)
T1ce = resample(T1ce, spacing=spacing, size=input_size)
T2 = resample(T2, spacing=spacing, size=input_size)
FLAIR = resample(FLAIR, spacing=spacing, size=input_size)
# convert to one batch of ndarray
T1 = sitk.GetArrayFromImage(T1).astype(np.float32)
T1ce = sitk.GetArrayFromImage(T1ce).astype(np.float32)
T2 = sitk.GetArrayFromImage(T2).astype(np.float32)
FLAIR = sitk.GetArrayFromImage(FLAIR).astype(np.float32)
image = np.stack((T1, T1ce, T2, FLAIR), 0)
# tensor
data = torch.from_numpy(image)
return data, dataID