Chris Oswald
commited on
Commit
•
f7a1cfb
1
Parent(s):
61fd3fa
added patient id splits
Browse files
SPIDER.py
CHANGED
@@ -23,7 +23,7 @@ from typing import Dict, List, Optional, Set
|
|
23 |
import numpy as np
|
24 |
|
25 |
import datasets
|
26 |
-
|
27 |
|
28 |
# TODO: Add BibTeX citation
|
29 |
# Find for instance the citation on arxiv or on the dataset repo/website
|
@@ -160,12 +160,19 @@ class SPIDER(datasets.GeneratorBasedBuilder):
|
|
160 |
# TODO: This method handles input defined in _split_generators to yield (key, example) tuples from the dataset.
|
161 |
# The `key` is for legacy reasons (tfds) and is not important in itself, but must be unique for each example.
|
162 |
|
163 |
-
#
|
|
|
164 |
np.random.seed(9999)
|
165 |
-
N_PATIENTS = 257
|
166 |
VALIDATE_SHARE = 0.3
|
167 |
TEST_SHARE = 0.2
|
168 |
TRAIN_SHARE = (1.0 - VALIDATE_SHARE - TEST_SHARE)
|
|
|
|
|
|
|
|
|
|
|
|
|
169 |
partition = np.random.choice(
|
170 |
['train', 'dev', 'test'],
|
171 |
p=[TRAIN_SHARE, VALIDATE_SHARE, TEST_SHARE],
|
@@ -176,9 +183,7 @@ class SPIDER(datasets.GeneratorBasedBuilder):
|
|
176 |
validate_ids = set(patient_ids[partition == 'dev'])
|
177 |
test_ids = set(patient_ids[partition == 'test'])
|
178 |
assert len(train_ids.union(validate_ids, test_ids)) == N_PATIENTS
|
179 |
-
|
180 |
-
|
181 |
-
|
182 |
# Import patient/scanner data and radiological gradings data
|
183 |
overview_data = import_csv_data(paths_dict['overview'])
|
184 |
grades_data = import_csv_data(paths_dict['gradings'])
|
@@ -196,17 +201,73 @@ class SPIDER(datasets.GeneratorBasedBuilder):
|
|
196 |
]
|
197 |
assert len(mask_files) > 0, "No mask files found--check directory path."
|
198 |
|
199 |
-
|
200 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
201 |
if split == 'train':
|
202 |
-
|
203 |
-
|
204 |
-
|
205 |
elif split == 'validate':
|
206 |
-
|
207 |
elif split == 'test':
|
|
|
|
|
|
|
|
|
|
|
208 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
209 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
210 |
def import_csv_data(filepath: str) -> List[Dict[str, str]]:
|
211 |
"""Import all rows of CSV file."""
|
212 |
results = []
|
|
|
23 |
import numpy as np
|
24 |
|
25 |
import datasets
|
26 |
+
import SimpleITK as sitk
|
27 |
|
28 |
# TODO: Add BibTeX citation
|
29 |
# Find for instance the citation on arxiv or on the dataset repo/website
|
|
|
160 |
# TODO: This method handles input defined in _split_generators to yield (key, example) tuples from the dataset.
|
161 |
# The `key` is for legacy reasons (tfds) and is not important in itself, but must be unique for each example.
|
162 |
|
163 |
+
# Config params
|
164 |
+
#TODO: make hardcoded values dynamic
|
165 |
np.random.seed(9999)
|
166 |
+
N_PATIENTS = 257
|
167 |
VALIDATE_SHARE = 0.3
|
168 |
TEST_SHARE = 0.2
|
169 |
TRAIN_SHARE = (1.0 - VALIDATE_SHARE - TEST_SHARE)
|
170 |
+
|
171 |
+
scan_types = ['t1', 't2', 't2_SPACE']
|
172 |
+
|
173 |
+
|
174 |
+
|
175 |
+
# Generate train/validate/test partitions of patient IDs
|
176 |
partition = np.random.choice(
|
177 |
['train', 'dev', 'test'],
|
178 |
p=[TRAIN_SHARE, VALIDATE_SHARE, TEST_SHARE],
|
|
|
183 |
validate_ids = set(patient_ids[partition == 'dev'])
|
184 |
test_ids = set(patient_ids[partition == 'test'])
|
185 |
assert len(train_ids.union(validate_ids, test_ids)) == N_PATIENTS
|
186 |
+
|
|
|
|
|
187 |
# Import patient/scanner data and radiological gradings data
|
188 |
overview_data = import_csv_data(paths_dict['overview'])
|
189 |
grades_data = import_csv_data(paths_dict['gradings'])
|
|
|
201 |
]
|
202 |
assert len(mask_files) > 0, "No mask files found--check directory path."
|
203 |
|
204 |
+
# Filter image and mask data based on scan types
|
205 |
+
image_files = [
|
206 |
+
file for file in image_files
|
207 |
+
if any(scan_type in file for scan_type in scan_types)
|
208 |
+
]
|
209 |
+
|
210 |
+
mask_files = [
|
211 |
+
file for file in mask_files
|
212 |
+
if any(scan_type in file for scan_type in scan_types)
|
213 |
+
]
|
214 |
+
|
215 |
+
# Subset train/validation/test partition images and mask files
|
216 |
if split == 'train':
|
217 |
+
subset_ids = train_ids
|
|
|
|
|
218 |
elif split == 'validate':
|
219 |
+
subset_ids = validate_ids
|
220 |
elif split == 'test':
|
221 |
+
subset_ids = test_ids
|
222 |
+
else:
|
223 |
+
subset_ids = None
|
224 |
+
raise ValueError(f'Split argument "{split}" is not recognized. \
|
225 |
+
Please enter one of ["train", "validate", "test"]')
|
226 |
|
227 |
+
image_files = [
|
228 |
+
file for file in image_files
|
229 |
+
if any(str(patient_id) in file.split('_')[0] for patient_id in subset_ids)
|
230 |
+
]
|
231 |
+
|
232 |
+
mask_files = [
|
233 |
+
file for file in mask_files
|
234 |
+
if any(str(patient_id) in file.split('_')[0] for patient_id in subset_ids)
|
235 |
+
]
|
236 |
+
assert len(image_files) == len(mask_files), "The number of image files\
|
237 |
+
does not match the number of mask files--verify subsetting operation."
|
238 |
+
|
239 |
+
# Shuffle order of patient scans
|
240 |
+
# (note that only images need to be shuffled since masks and metadata
|
241 |
+
# will be linked to the selected image)
|
242 |
+
np.random.shuffle(image_files)
|
243 |
+
|
244 |
+
## Generate next example
|
245 |
+
# ----------------------
|
246 |
+
for example in image_files:
|
247 |
+
|
248 |
+
# Extract linking data
|
249 |
+
scan_id = example.replace('.mha', '')
|
250 |
+
patient_id = scan_id.split('_')[0]
|
251 |
+
scan_type = '_'.join(scan_id.split('_')[1:])
|
252 |
+
|
253 |
+
# Load .mha file
|
254 |
+
image_path = os.path.join(paths_dict['images'], 'images', example)
|
255 |
+
image = sitk.ReadImage(image_path)
|
256 |
+
|
257 |
+
# Convert .mha image to numeric array
|
258 |
+
image_array = sitk.GetArrayFromImage(image)
|
259 |
|
260 |
+
# Extract overview data corresponding to image
|
261 |
+
|
262 |
+
# Extract patient radiological gradings corresponding to image
|
263 |
+
|
264 |
+
|
265 |
+
|
266 |
+
|
267 |
+
|
268 |
+
|
269 |
+
|
270 |
+
|
271 |
def import_csv_data(filepath: str) -> List[Dict[str, str]]:
|
272 |
"""Import all rows of CSV file."""
|
273 |
results = []
|