forked from B1tB3ast/CheXpred
-
Notifications
You must be signed in to change notification settings - Fork 0
Expand file tree
/
Copy pathdata_process.py
More file actions
148 lines (128 loc) · 4.77 KB
/
data_process.py
File metadata and controls
148 lines (128 loc) · 4.77 KB
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
import os
import glob
import numpy as np
import pandas as pd
import csv
import matplotlib.pyplot as plt
from tqdm import tqdm
from PIL import Image
import h5py
import cv2
from typing import *
from pathlib import Path
import torch
from torchvision.transforms import Compose, Resize, CenterCrop, ToTensor, Normalize
def load_data(filepath):
dataframe = pd.read_csv(filepath)
return dataframe
def get_cxr_paths_list(filepath):
dataframe = load_data(filepath)
cxr_paths = dataframe['Path']
return cxr_paths
'''
This function resizes and zero pads image
'''
def preprocess(img, desired_size=320):
old_size = img.size
ratio = float(desired_size)/max(old_size)
new_size = tuple([int(x*ratio) for x in old_size])
img = img.resize(new_size, Image.ANTIALIAS)
# create a new image and paste the resized on it
new_img = Image.new('L', (desired_size, desired_size))
new_img.paste(img, ((desired_size-new_size[0])//2,
(desired_size-new_size[1])//2))
return new_img
def img_to_hdf5(cxr_paths: List[Union[str, Path]], out_filepath: str, resolution=320):
"""
Convert directory of images into a .h5 file given paths to all
images.
"""
dset_size = len(cxr_paths)
failed_images = []
with h5py.File(out_filepath,'w') as h5f:
img_dset = h5f.create_dataset('cxr', shape=(dset_size, resolution, resolution))
for idx, path in enumerate(tqdm(cxr_paths)):
try:
# read image using cv2
img = cv2.imread(str(path))
# convert to PIL Image object
img = cv2.cvtColor(img, cv2.COLOR_BGR2RGB)
img_pil = Image.fromarray(img)
# preprocess
img = preprocess(img_pil, desired_size=resolution)
img_dset[idx] = img
except Exception as e:
failed_images.append((path, e))
print(f"{len(failed_images)} / {len(cxr_paths)} images failed to be added to h5.", failed_images)
def get_files(directory):
files = []
for (dirpath, dirnames, filenames) in os.walk(directory):
for file in filenames:
if file.endswith(".jpg"):
files.append(os.path.join(dirpath, file))
return files
def get_cxr_path_csv(out_filepath, directory):
files = get_files(directory)
file_dict = {"Path": files}
df = pd.DataFrame(file_dict)
df.to_csv(out_filepath, index=False)
def section_start(lines, section=' IMPRESSION'):
for idx, line in enumerate(lines):
if line.startswith(section):
return idx
return -1
def section_end(lines, section_start):
num_lines = len(lines)
def getIndexOfLast(l, element):
""" Get index of last occurence of element
@param l (list): list of elements
@param element (string): element to search for
@returns (int): index of last occurrence of element
"""
i = max(loc for loc, val in enumerate(l) if val == element)
return i
def write_report_csv(cxr_paths, txt_folder, out_path):
imps = {"filename": [], "impression": []}
txt_reports = []
for cxr_path in cxr_paths:
tokens = cxr_path.split('/')
study_num = tokens[-2]
patient_num = tokens[-3]
patient_group = tokens[-4]
txt_report = txt_folder + patient_group + '/' + patient_num + '/' + study_num + '.txt'
filename = study_num + '.txt'
f = open(txt_report, 'r')
s = f.read()
s_split = s.split()
if "IMPRESSION:" in s_split:
begin = getIndexOfLast(s_split, "IMPRESSION:") + 1
end = None
end_cand1 = None
end_cand2 = None
# remove recommendation(s) and notification
if "RECOMMENDATION(S):" in s_split:
end_cand1 = s_split.index("RECOMMENDATION(S):")
elif "RECOMMENDATION:" in s_split:
end_cand1 = s_split.index("RECOMMENDATION:")
elif "RECOMMENDATIONS:" in s_split:
end_cand1 = s_split.index("RECOMMENDATIONS:")
if "NOTIFICATION:" in s_split:
end_cand2 = s_split.index("NOTIFICATION:")
elif "NOTIFICATIONS:" in s_split:
end_cand2 = s_split.index("NOTIFICATIONS:")
if end_cand1 and end_cand2:
end = min(end_cand1, end_cand2)
elif end_cand1:
end = end_cand1
elif end_cand2:
end = end_cand2
if end == None:
imp = " ".join(s_split[begin:])
else:
imp = " ".join(s_split[begin:end])
else:
imp = 'NO IMPRESSION'
imps["impression"].append(imp)
imps["filename"].append(filename)
df = pd.DataFrame(data=imps)
df.to_csv(out_path, index=False)