-
Notifications
You must be signed in to change notification settings - Fork 2
/
preprocessing.py
178 lines (146 loc) · 5.87 KB
/
preprocessing.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
165
166
167
168
169
170
171
172
173
174
175
176
177
178
# -*- coding: utf-8 -*-
"""
Created on Fri Apr 2 20:17:22 2021
functions used for downsampling
@author: klin0
"""
import os
import numpy as np
import nibabel as nib
from scipy import ndimage
class Param():
'''
parameter class to store all the parameters
'''
def __init__(self, resize_option = "by_zdist"):
self.window_min = -100
self.window_max = 400
self.patch_shape = (128, 128, 16) # used for resizing by slice spacing
self.equalize_histogram = False
self.normalize = True
self.resize_option = resize_option # options are "by_zdist" or "by_vol"
self.zoom_order = 3
if self.resize_option == "by_zdist":
self.zdist = 2 # set z spacing to zdist mm, only vlid when resize option is by zdist
elif self.resize_option == "by_vol":
self.resized_vol_shape = (128, 128, 128) # used for resizing volume to certain shape
else:
raise ValueError(f"{self.resize_option} is not a valid resize option")
self.output_type = 'npy'
# preprocessing functions
def read_nii(f):
img_obj = nib.load(f)
img_data = img_obj.get_fdata()
return img_data, img_obj.header
def hist_eq(image, number_bins=32):
# histogram equalization
# adopt from http://www.janeriksolem.net/2009/06/histogram-equalization-with-python-and.html
# get image histogram
image_histogram, bins = np.histogram(image.flatten(), number_bins, density=True)
cdf = image_histogram.cumsum() # cumulative distribution function
cdf = 255 * cdf / cdf[-1] # normalize
# use linear interpolation of cdf to find new pixel values
image_equalized = np.interp(image.flatten(), bins[:-1], cdf)
return image_equalized.reshape(image.shape)
def windowing(nparray_2d, _min, _max):
# Setting hounsfield unit values to [−100, 400] to discard irrelevant structures
np.clip(nparray_2d, _min, _max)
def norm(nparray):
# normalize scans to [0,1]
_min = nparray.min()
_max = nparray.max()
nparray = nparray - _min
nparray = nparray / (_max - _min)
return nparray
def norm_zscore(nparray):
# normalize 2d scands by mean and standard deviation
mean = nparray.mean()
std = nparray.std()
nparray = nparray - mean
nparray /= std
return nparray
def resize_volume(orig_volume, zdist, params, order):
"""
resize orig_volume to desired dimension defined in parameters
zdist: zdist of the volume
"""
resize_factor = [0]*3
if params.resize_option == "by_zdist":
for i in range(2):
resize_factor[i] = params.patch_shape[i]/orig_volume.shape[i]
# rescale scan spacing to 2mm
resize_factor[2] = zdist/params.zdist
elif params.resize_option == "by_vol":
for i in range(3):
resize_factor[i] = params.resized_vol_shape[i]/orig_volume.shape[i]
else:
raise ValueError(f"{params.resize_option} is not a valid resize option")
resized_vol = ndimage.zoom(orig_volume, resize_factor, order = order)
return resized_vol
def upsample_volume(downsampled_vol, original_vol_shape, order):
resize_factor = [0]*3
for i in range(3):
resize_factor[i] = original_vol_shape[i]/downsampled_vol.shape[i]
return ndimage.zoom(downsampled_vol, resize_factor, order = order)
def preprocessing_vol(f_vol, param):
vol, header = read_nii(f_vol)
zdist = header['srow_z'][-2]
# windowing
vol = np.clip(vol, param.window_min, param.window_max)
# resizing vol
vol = resize_volume(vol, zdist, param, order = param.zoom_order)
# histogram equalization
if param.equalize_histogram:
for i in range(vol.shape[-1]):
vol[:,:,i] = hist_eq(vol[:,:,i])
# normalizing
if param.normalize:
vol = norm(vol)
# vol = norm_zscore(vol)
# output zdist for preprocessing_mask (spacing between scan and mask is different for some cases)
return vol, zdist
def preprocessing_mask(f_mask, zdist, param):
mask, _ = read_nii(f_mask)
mask = resize_volume(mask, zdist, param, order = param.zoom_order)
mask = np.rint(mask)
return mask.astype(int)
def load_filepaths_to_dictionaries(path):
"""
output dictionaries that record the input file paths
"""
volumes = {}
segments = {}
for dirname, _, filenames in os.walk(path):
for filename in filenames:
if filename.startswith('volume-'):
num = filename.split('-')[1]
num = num.split('.')[0]
volumes[int(num)] = os.path.join(dirname, filename)
elif filename.startswith('segmentation-'):
num = filename.split('-')[1]
num = num.split('.')[0]
segments[int(num)] = os.path.join(dirname, filename)
assert(len(volumes.keys()) == len(segments.keys()))
for k in volumes.keys():
assert(k in segments)
return volumes, segments
if __name__ == "__main__":
path = r'kaggle/input'
params = Param(resize_option = "by_vol")
volume_dict, segment_dict = load_filepaths_to_dictionaries(path = path)
for key in volume_dict.keys():
vol, zdist = preprocessing_vol(volume_dict[key], params)
mask = preprocessing_mask(segment_dict[key], zdist, params)
if vol.shape != mask.shape:
print("key, vol, mask: ", str(key), vol.shape, mask.shape)
np.save('vol'+ str(key)+'.npy', vol)
np.save('mask'+ str(key)+'.npy', mask)
# visualization
# upsample vol and mask
k=10
orig_vol, _ = read_nii(volume_dict[key])
orig_mask, _ = read_nii(segment_dict[key])
up_vol = upsample_volume(vol, orig_vol.shape, order = 3)
up_mask = upsample_volume(mask, orig_vol.shape, order = 3)
import lits_util
lits_util.plot_mask_comparison_over_vol(up_vol, orig_mask, orig_mask+up_mask, index_start = 46, step = 4)