-
Notifications
You must be signed in to change notification settings - Fork 7
/
prepare_partitions.py
1229 lines (923 loc) · 49.9 KB
/
prepare_partitions.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
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
530
531
532
533
534
535
536
537
538
539
540
541
542
543
544
545
546
547
548
549
550
551
552
553
554
555
556
557
558
559
560
561
562
563
564
565
566
567
568
569
570
571
572
573
574
575
576
577
578
579
580
581
582
583
584
585
586
587
588
589
590
591
592
593
594
595
596
597
598
599
600
601
602
603
604
605
606
607
608
609
610
611
612
613
614
615
616
617
618
619
620
621
622
623
624
625
626
627
628
629
630
631
632
633
634
635
636
637
638
639
640
641
642
643
644
645
646
647
648
649
650
651
652
653
654
655
656
657
658
659
660
661
662
663
664
665
666
667
668
669
670
671
672
673
674
675
676
677
678
679
680
681
682
683
684
685
686
687
688
689
690
691
692
693
694
695
696
697
698
699
700
701
702
703
704
705
706
707
708
709
710
711
712
713
714
715
716
717
718
719
720
721
722
723
724
725
726
727
728
729
730
731
732
733
734
735
736
737
738
739
740
741
742
743
744
745
746
747
748
749
750
751
752
753
754
755
756
757
758
759
760
761
762
763
764
765
766
767
768
769
770
771
772
773
774
775
776
777
778
779
780
781
782
783
784
785
786
787
788
789
790
791
792
793
794
795
796
797
798
799
800
801
802
803
804
805
806
807
808
809
810
811
812
813
814
815
816
817
818
819
820
821
822
823
824
825
826
827
828
829
830
831
832
833
834
835
836
837
838
839
840
841
842
843
844
845
846
847
848
849
850
851
852
853
854
855
856
857
858
859
860
861
862
863
864
865
866
867
868
869
870
871
872
873
874
875
876
877
878
879
880
881
882
883
884
885
886
887
888
889
890
891
892
893
894
895
896
897
898
899
900
901
902
903
904
905
906
907
908
909
910
911
912
913
914
915
916
917
918
919
920
921
922
923
924
925
926
927
928
929
930
931
932
933
934
935
936
937
938
939
940
941
942
943
944
945
946
947
948
949
950
951
952
953
954
955
956
957
958
959
960
961
962
963
964
965
966
967
968
969
970
971
972
973
974
975
976
977
978
979
980
981
982
983
984
985
986
987
988
989
990
991
992
993
994
995
996
997
998
999
1000
"""
This script contains the code used to obtain the
assembly data dataframes with text and categorical
labels, partitioned for train/evaluate.
Also, we prepare here the dataframes for prompts
(zero-shot) and transferability experiments.
"""
import os
import json
import glob
import pandas as pd
import numpy as np
from local_data.constants import *
if not os.path.exists(PATH_DATAFRAME_PRETRAIN):
os.mkdir(PATH_DATAFRAME_PRETRAIN)
if not os.path.exists(PATH_DATAFRAME_TRANSFERABILITY):
os.mkdir(PATH_DATAFRAME_TRANSFERABILITY)
if not os.path.exists(PATH_DATAFRAME_TRANSFERABILITY_CLASSIFICATION):
os.mkdir(PATH_DATAFRAME_TRANSFERABILITY_CLASSIFICATION)
if not os.path.exists(PATH_DATAFRAME_TRANSFERABILITY_SEGMENTATION):
os.mkdir(PATH_DATAFRAME_TRANSFERABILITY_SEGMENTATION)
def adequate_01_eyepacs():
labels_dr = {0: "no diabetic retinopathy", 1: "mild diabetic retinopathy", 2: "moderate diabetic retinopathy",
3: "severe diabetic retinopathy", 4: "proliferative diabetic retinopathy"}
path_dataset = "01_EYEPACS/"
partitions = ["train", "test", "val"]
data = []
for iPartition in partitions:
print(iPartition)
dataframe = pd.read_csv(PATH_DATASETS + path_dataset + iPartition + ".csv")
for iFile in range(dataframe.shape[0]):
print(iFile, end="\r")
image_path = path_dataset + "documents/" + dataframe["image_id"][iFile].split("/")[-1].replace(".jpg", ".jpeg")
categories, atributes = [], []
categories.append(labels_dr[dataframe["dr"][iFile]])
if os.path.isfile(PATH_DATASETS + image_path):
data.append({"image": image_path,
"atributes": atributes,
"categories": categories})
df_out = pd.DataFrame(data)
df_out.to_csv(PATH_DATAFRAME_PRETRAIN + "01_EYEPACS.csv")
def adequate_02_messidor():
labels_dr = {0: "no diabetic retinopathy", 1: "mild diabetic retinopathy", 2: "moderate diabetic retinopathy",
3: "severe diabetic retinopathy", 4: "proliferative diabetic retinopathy"}
labels_dme = {0: "no diabetic macular edema", 1: "diabetic macular edema"}
labels_gradable = {0: "noisy", 1: "clean"}
path_dataset = "02_MESSIDOR/"
# The link to MESSIDOR2 labels is at:
# https://www.kaggle.com/datasets/google-brain/messidor2-dr-grades?select=messidor_data.csv
dataframe = pd.read_csv(PATH_DATASETS + path_dataset + "messidor_data.csv")
data = []
for iFile in range(dataframe.shape[0]):
image_path = path_dataset + "documents/" + dataframe["image_id"][iFile].replace(".jpg", ".JPG")
categories, atributes = [], []
# Noisy/Clean
atributes.append(labels_gradable[dataframe["adjudicated_gradable"][iFile]])
if dataframe["adjudicated_gradable"][iFile] == 1:
categories.append(labels_dr[dataframe["adjudicated_dr_grade"][iFile]])
categories.append(labels_dme[dataframe["adjudicated_dme"][iFile]])
if os.path.isfile(PATH_DATASETS + image_path):
if len(categories) > 0:
data.append({"image": image_path,
"atributes": atributes,
"categories": categories})
df_out = pd.DataFrame(data)
df_out.to_csv(PATH_DATAFRAME_TRANSFERABILITY_CLASSIFICATION + "02_MESSIDOR.csv")
def adequate_03_idrid():
path_dataset = "03_IDRID/"
data = []
# A.Segmentation
subpath = "A.%20Segmentation/A. Segmentation/"
subpath_images = "1. Original Images/a. Training Set/"
subpath_gt = "2. All Segmentation Groundtruths/a. Training Set/"
annotations_paths = ["1. Microaneurysms", "2. Haemorrhages", "3. Hard Exudates", "4. Soft Exudates"]
annotations_abbreviations = ["MA", "HE", "EX", "SE"]
annotations_categories = ["microaneurysms", "haemorrhages", "hard exudates", "soft exudates"]
files_segmentation = os.listdir(PATH_DATASETS + path_dataset + subpath + subpath_images)
for iFile in files_segmentation:
image_path = path_dataset + subpath + subpath_images + iFile
categories = []
atributes = []
for i in range(len(annotations_categories)):
annotation_path = PATH_DATASETS + path_dataset + subpath + subpath_gt + annotations_paths[i] + "/"\
+ iFile.replace(".jpg","_" + annotations_abbreviations[i] + ".tif")
if os.path.isfile(annotation_path):
categories.append(annotations_categories[i])
data.append({"image": image_path,
"atributes": atributes,
"categories": categories})
# B.Grading
labels_dr = {0: "no diabetic retinopathy", 1: "mild diabetic retinopathy", 2: "moderate diabetic retinopathy",
3: "severe diabetic retinopathy", 4: "proliferative diabetic retinopathy"}
labels_dme = {0: "no referable diabetic macular edema", 1: "non clinically significant diabetic macular edema",
2: "diabetic macular edema"}
subpath = "B.%20Disease%20Grading/B. Disease Grading/"
subpath_images = "1. Original Images/a. Training Set/"
dataframe = "2. Groundtruths/a. IDRiD_Disease Grading_Training Labels.csv"
dataframe = pd.read_csv(PATH_DATASETS + path_dataset + subpath + dataframe)
for iFile in range(dataframe.shape[0]):
image_path = path_dataset + subpath + subpath_images + dataframe["Image name"][iFile] + ".jpg"
categories = []
atributes = []
categories.append(labels_dr[dataframe["Retinopathy grade"][iFile]])
categories.append(labels_dme[dataframe["Risk of macular edema "][iFile]])
if os.path.isfile(PATH_DATASETS + image_path):
data.append({"image": image_path,
"atributes": atributes,
"categories": categories})
df_out = pd.DataFrame(data)
df_out.to_csv(PATH_DATAFRAME_PRETRAIN + "03_IDRID.csv")
def adequate_03_idrid_segmentation():
path_dataset = "03_IDRID/"
data = []
# A.Segmentation
subpath = "A.%20Segmentation/A. Segmentation/"
subpath_images = "1. Original Images/"
subpath_masks = "2. All Segmentation Groundtruths/"
partitions = ["a. Training Set/", "b. Testing Set/"]
partitions_names = ["train", "test"]
categories_paths = ["1. Microaneurysms/", "2. Haemorrhages/", "3. Hard Exudates/", "4. Soft Exudates/"]
annotations_abbreviations = ["MA", "HE", "EX", "SE"]
annotations_categories = ["microaneurysms", "haemorrhages", "hard_exudates", "soft_exudates"]
for iPartition in range(len(partitions)):
for iCategory in range(len(categories_paths)):
data = []
files = os.listdir(PATH_DATASETS + path_dataset + subpath + subpath_masks + partitions[iPartition] + categories_paths[iCategory])
for iFile in files:
image_path = path_dataset + subpath + subpath_images +\
partitions[iPartition] + iFile.replace(".tif", ".jpg").replace("_" + annotations_abbreviations[iCategory], "")
mask_path = path_dataset + subpath + subpath_masks + partitions[iPartition] + categories_paths[iCategory] + iFile
data.append({"image": image_path,
"mask": mask_path})
df_out = pd.DataFrame(data)
df_out.to_csv(PATH_DATAFRAME_TRANSFERABILITY_SEGMENTATION + "03_IDRID_" +
annotations_categories[iCategory] + '_' + partitions_names[iPartition] + '.csv')
def adequate_04_rfmid():
template_diseases = {'DR': 'diabetic retinopathy', 'ARMD': 'age-related macular degeneration',
'MH': 'media haze', 'DN': 'drusens', 'MYA': 'myopia', 'BRVO': 'branch retinal vein occlusion',
'TSLN': 'tessellation', 'ERM': 'epiretinal membrane', 'LS': 'laser scar',
'MS': 'macular scar', 'CSR': 'central serous retinopathy', 'ODC': 'optic disc cupping',
'CRVO': 'central retinal vein occlusion', 'TV': 'tortuous vessels', 'AH': 'asteroid hyalosis',
'ODP': 'optic disc pallor', 'ODE': 'optic disc edema', 'ST': 'shunt',
'AION': 'anterior ischemic optic neuropathy', 'PT': 'parafoveal telangiectasia',
'RT': 'retinal traction', 'RS': 'retinitis', 'CRS': 'chorioretinitis', 'EDN': 'edudates',
'RPEC': 'retinal pigment epithelium changes', 'MHL': 'macular hole',
'RP': 'retinitis pigmentosa', 'CWS':'cotton wool spots', 'CB': 'colobomas',
'ODPM': 'optic disc pit maculopathy', 'PRH': 'preretinal haemorrhage',
'MNF': 'myelinated nerve fibers', 'HR': 'haemorrhagic retinopathy',
'CRAO': 'central retinal artery occlusion', 'TD': 'tilted disc',
'CME': 'cystoid macular edema', 'PTCR': 'post traumatic choroidal rupture',
'CF': 'choroidal folds', 'VH': 'vitreous haemorrhage', 'MCA': 'macroaneurysm',
'VS': 'vasculitis', 'BRAO': 'branch retinal artery occlusion', 'PLQ': 'plaque',
'HPED': 'haemorrhagic pigment epithelial detachment', 'CL': 'collaterals'}
path_dataset = "04_RFMid/"
partitions = ["Training", "Validation", "Testing"]
letters = ["a", "b", "c"]
data = []
for iPartition in range(len(partitions)):
subpath_images = "1. Original Images/" + letters[iPartition] + ". " + partitions[iPartition] + " Set/"
subpath_dataframe = "2. Groundtruths/" + letters[iPartition] + ". RFMiD_" + partitions[iPartition] + "_Labels.csv"
dataframe = pd.read_csv(PATH_DATASETS + path_dataset + subpath_dataframe)
for iFile in range(dataframe.shape[0]):
image_path = path_dataset + subpath_images + str(dataframe["ID"][iFile]) + ".png"
categories, atributes = [], []
if dataframe["Disease_Risk"][iFile] == 1:
categories.append("a disease")
ids = np.argwhere(np.array(dataframe)[iFile, 2:])
for i in list(ids):
dis_abreviation = dataframe.columns.to_list()[i[0]+2]
categories.append(template_diseases[dis_abreviation])
else:
categories.append("no disease")
categories.append("healthy")
if os.path.isfile(PATH_DATASETS + image_path):
data.append({"image": image_path,
"atributes": atributes,
"categories": categories})
df_out = pd.DataFrame(data)
df_out.to_csv(PATH_DATAFRAME_PRETRAIN + "04_RFMid.csv")
def adequate_05_1000x39():
categories_template = {'0.0.Normal': 'normal', '0.1.Tessellated fundus': 'tessellation',
'0.2.Large optic cup': 'large optic cup', '0.3.DR1': 'mild diabetic retinopathy',
'1.0.DR2': 'moderate diabetic retinopathy', '1.1.DR3': 'severe diabetic retinopathy',
'2.0.BRVO': 'branch retinal vein occlusion', '2.1.CRVO': 'central retinal vein occlusion',
'3.RAO': 'central retinal artery occlusion', '4.Rhegmatogenous RD': 'retina detachment',
'5.0.CSCR':'central serous retinopathy', '5.1.VKH disease': 'Vogt-Koyanagi syndrome',
'6.Maculopathy': 'maculopathy', '7.ERM': 'epiretinal membrane', '8.MH': 'macular hole',
'9.Pathological myopia': 'pathologic myopia', '10.0.Possible glaucoma': 'glaucoma',
'10.1.Optic atrophy': 'optic atrophy',
'11.Severe hypertensive retinopathy': 'severe hypertensive retinopathy',
'12.Disc swelling and elevation': 'disc swelling and elevation',
'13.Dragged Disc': 'dragged disk',
'14.Congenital disc abnormality': 'congenital disk abnormality',
'15.0.Retinitis pigmentosa':'retinitis pigmentosa',
'15.1.Bietti crystalline dystrophy': 'Bietti crystalline dystrophy',
'16.Peripheral retinal degeneration and break': 'peripheral retinal degeneration and break',
'17.Myelinated nerve fiber': 'myelinated nerve fibers',
'18.Vitreous particles': 'vitreous haemorrhage', '19.Fundus neoplasm': 'neoplasm',
'20.Massive hard exudates': 'hard exudates',
'21.Yellow-white spots-flecks': 'yellow-white spots flecks',
'22.Cotton-wool spots':'cotton wool spots', '23.Vessel tortuosity':'tortuous vessels',
'24.Chorioretinal atrophy-coloboma': 'colobomas',
'25.Preretinal hemorrhage': 'preretinal haemorrhage', '26.Fibrosis': 'fibrosis',
'27.Laser Spots': 'laser scar', '28.Silicon oil in eye': 'silicon oil',
'29.0.Blur fundus without PDR': 'no proliferative diabetic retinopathy',
'29.1.Blur fundus with suspected PDR': 'proliferative diabetic retinopathy'}
categories_test = ["0.0.Normal", "15.0.Retinitis pigmentosa", "8.MH"]
path_dataset = "05_1000x39/"
data = []
categories = os.listdir(PATH_DATASETS + path_dataset)
[categories.remove(iCategory) for iCategory in categories_test]
for iCategory in categories:
images = os.listdir(PATH_DATASETS + path_dataset + iCategory + "/")
for iImage in images:
data.append({"image": path_dataset + iCategory + "/" + iImage,
"atributes": [],
"categories": [categories_template[iCategory]]})
df_out = pd.DataFrame(data)
df_out.to_csv(PATH_DATAFRAME_PRETRAIN + "05_1000x39.csv")
data = []
for iCategory in categories_test:
images = os.listdir(PATH_DATASETS + path_dataset + iCategory + "/")
images = images[0:20]
for iImage in images:
data.append({"image": path_dataset + iCategory + "/" + iImage,
"atributes": [],
"categories": [categories_template[iCategory]]})
df_out = pd.DataFrame(data)
df_out.to_csv(PATH_DATAFRAME_TRANSFERABILITY_CLASSIFICATION + "05_20x4.csv")
def adequate_06_DEN():
path_dataset = "06_DEN/"
data = []
partitions = ["DeepEyeNet_train.json", "DeepEyeNet_test.json", "DeepEyeNet_valid.json"]
for iPartition in partitions:
f = open(PATH_DATASETS + path_dataset + iPartition)
meta = json.load(f)
for iSample in meta:
image_path = path_dataset + list(iSample.keys())[0]
categories, atributes = [], []
info = iSample[list(iSample.keys())[0]]
categories.extend(info["keywords"].split(", "))
categories.extend(info["clinical-description"].split(". "))
if "" in categories:
categories.remove("")
if os.path.isfile(PATH_DATASETS + image_path):
data.append({"image": image_path,
"atributes": atributes,
"categories": categories})
df_out = pd.DataFrame(data)
df_out.to_csv(PATH_DATAFRAME_PRETRAIN + "06_DEN.csv")
def adequate_07_lag():
path_dataset = "07_LAG/"
categories_paths = ["non_glaucoma", "suspicious_glaucoma"]
categories = ["no glaucoma", "glaucoma"]
data = []
for i in range(len(categories_paths)):
images = os.listdir(PATH_DATASETS + path_dataset + categories_paths[i] + "/image/")
for iImage in images:
data.append({"image": path_dataset + categories_paths[i] + "/image/" + iImage,
"atributes": [],
"categories": [categories[i]]})
df_out = pd.DataFrame(data)
df_out.to_csv(PATH_DATAFRAME_PRETRAIN + "07_LAG.csv")
def adequate_08_odir5k():
path_dataset = "08_ODIR-5K/"
dataframe = pd.read_csv(PATH_DATASETS + path_dataset + "full_df.csv")
# Train and Incremental test division
mask_cat = np.logical_and(dataframe["C"].values == 1, dataframe[["D", "G", "C", "A", "H", "M", "O"]].values.sum(-1) == 1)
mask_myo = np.logical_and(dataframe["M"].values == 1, dataframe[["D", "G", "C", "A", "H", "M", "O"]].values.sum(-1) == 1)
mask_normal = np.logical_and(dataframe["N"].values == 1, dataframe[["D", "G", "C", "A", "H", "M", "O"]].values.sum(-1) == 0)
mask_normal[np.argwhere(mask_normal == 1)[200:]] = False
mask_test = np.logical_or(mask_cat, mask_myo)
mask_test = np.logical_or(mask_test, mask_normal)
dataframe_train = dataframe[np.logical_not(mask_test)]
dataframe_test = dataframe[mask_test]
# Train subset
data = []
for iFile in range(dataframe_train.shape[0]):
id = dataframe_train["ID"].values[iFile]
for iEye in ["Right", "Left"]:
image_path = path_dataset + "preprocessed_images/" + str(id) + "_" + (iEye).lower() + ".jpg"
categories = []
description = dataframe_train[(iEye + "-Diagnostic Keywords")].values[iFile]
if "myop" not in description and "cataract" not in description:
categories.extend(description.split("，"))
if os.path.isfile(PATH_DATASETS + image_path):
data.append({"image": image_path,
"atributes": [],
"categories": categories})
df_out = pd.DataFrame(data)
df_out.to_csv(PATH_DATAFRAME_PRETRAIN + "08_ODIR.csv")
# Test subset
data = []
counter_n, counter_m, counter_c = 1, 1, 1
for iFile in range(dataframe_test.shape[0]):
id = dataframe_test["ID"].values[iFile]
for iEye in ["Right", "Left"]:
image_path = path_dataset + "preprocessed_images/" + str(id) + "_" + (iEye).lower() + ".jpg"
description = dataframe_test[(iEye + "-Diagnostic Keywords")].values[iFile]
if "myop" in description and counter_m<=200:
if os.path.isfile(PATH_DATASETS + image_path):
data.append({"image": image_path,
"atributes": [],
"categories": ["pathologic myopia"]})
counter_m += 1
if "cataract" in description and counter_c<=200:
if os.path.isfile(PATH_DATASETS + image_path):
data.append({"image": image_path,
"atributes": [],
"categories": ["cataract"]})
counter_c += 1
if "normal" in description and counter_n<=200:
if os.path.isfile(PATH_DATASETS + image_path):
data.append({"image": image_path,
"atributes": [],
"categories": ["normal"]})
counter_n += 1
df_out = pd.DataFrame(data)
df_out.to_csv(PATH_DATAFRAME_TRANSFERABILITY_CLASSIFICATION + "08_ODIR200x3.csv")
def adequate_09_papila():
path_dataset = "09_PAPILA/"
subpath_images = "FundusImages/"
dataframes = [pd.read_excel(PATH_DATASETS + path_dataset + "ClinicalData/patient_data_od.xlsx"),
pd.read_excel(PATH_DATASETS + path_dataset + "ClinicalData/patient_data_os.xlsx")]
labels_glaucoma = {0: "normal", 1: "glaucoma", 2: "glaucoma"}
data = []
for iFile in range(dataframes[0].shape[0]-2):
id = dataframes[0]["Unnamed: 0"][iFile+2][1:]
image_path = path_dataset + subpath_images + "RET" + id + "OD" + ".jpg"
if os.path.isfile(PATH_DATASETS + image_path):
data.append({"image": image_path,
"atributes": [],
"categories": [labels_glaucoma[dataframes[0]["Diagnosis"][iFile+2]]]})
image_path = path_dataset + subpath_images + "RET" + id + "OS" + ".jpg"
if os.path.isfile(PATH_DATASETS + image_path):
data.append({"image": image_path,
"atributes": [],
"categories": [labels_glaucoma[dataframes[1]["Diagnosis"][iFile+2]]]})
df_out = pd.DataFrame(data)
df_out.to_csv(PATH_DATAFRAME_PRETRAIN + "09_PAPILA.csv")
def adequate_10_paraguay():
path_dataset = "10_PARAGUAY/"
data = []
subpaths = ["1. No DR signs", "2. Mild (or early) NPDR", "3. Moderate NPDR",
"4. Severe NPDR", "5. Very Severe NPDR", "6. PDR", "7. Advanced PDR"]
categories = ["no diabetic retinopathy", "mild diabetic retinopathy",
"moderate diabetic retinopathy", "severe diabetic retinopathy",
"severe diabetic retinopathy", "proliferative diabetic retinopathy",
"proliferative diabetic retinopathy"
]
for iPath in range(len(subpaths)):
images = os.listdir(PATH_DATASETS + path_dataset + subpaths[iPath] + "/")
for iImage in images:
data.append({"image": path_dataset + subpaths[iPath] + "/" + iImage,
"atributes": [],
"categories": [categories[iPath]]})
df_out = pd.DataFrame(data)
df_out.to_csv(PATH_DATAFRAME_PRETRAIN + "10_PARAGUAY.csv")
def adequate_11_stare():
path_dataset = "11_STARE/"
data = []
metadata = "all-mg-codes.txt"
for line in open(PATH_DATASETS + path_dataset + metadata):
categories, atributes = [], []
columns = line.strip().split("\t")
image_path = path_dataset + "documents/" + columns[0] + ".ppm"
description = columns[-1].split("\n")[0].lower().split(" ")[-1].replace("\"", "")
categories.extend(description.replace("\t", "").replace(" and ", " or ").replace("?", "").split(" or "))
if os.path.isfile(PATH_DATASETS + image_path):
data.append({"image": image_path,
"atributes": [],
"categories": categories})
df_out = pd.DataFrame(data)
df_out.to_csv(PATH_DATAFRAME_PRETRAIN + "11_STARE.csv")
def adequate_12_aria():
path_dataset = "12_ARIA/"
categories_subpath = ["aria_a_markups", "aria_c_markups", "aria_d_markups"]
categories = ["age-related macular degeneration", "normal", "diabetic retinopathy"]
data = []
for i in range(len(categories)):
for iFile in os.listdir(PATH_DATASETS + path_dataset + categories_subpath[i] + "/"):
if iFile != "Thumbs.db":
data.append({"image": path_dataset + categories_subpath[i] + "/" + iFile,
"atributes": [],
"categories": [categories[i]]})
df_out = pd.DataFrame(data)
df_out.to_csv(PATH_DATAFRAME_PRETRAIN + "12_ARIA.csv")
def adequate_13_fives():
path_dataset = "13_FIVES/"
images_subpath = ["train/Original/", "test/Original/"]
labels_dme = {"A": "age related macular degeneration",
"D": "diabetic retinopathy",
"G": "glaucoma",
"N": "normal"}
data = []
for iSubpath in images_subpath:
files = os.listdir(PATH_DATASETS + path_dataset + iSubpath)
for iFile in files:
if iFile != "Thumbs.db":
category__code = iFile.split(".")[0].split("_")[-1]
data.append({"image": path_dataset + iSubpath + iFile,
"atributes": [],
"categories": [labels_dme[category__code]]})
df_out = pd.DataFrame(data)
df_out.to_csv(PATH_DATAFRAME_TRANSFERABILITY_CLASSIFICATION + "13_FIVES.csv")
def adequate_14_agar300():
path_dataset = "14_AGAR300/"
finding = ["microaneurysms", "diabetic retinopathy"]
images_subpath = "img/"
data = []
files = os.listdir(PATH_DATASETS + path_dataset + images_subpath)
for iFile in files:
data.append({"image": path_dataset + images_subpath + iFile,
"atributes": [],
"categories": finding})
df_out = pd.DataFrame(data)
df_out.to_csv(PATH_DATAFRAME_PRETRAIN + "14_AGAR300.csv")
def adequate_15_aptos():
path_dataset = "15_APTOS/"
labels_dr = {0: "no diabetic retinopathy", 1: "mild diabetic retinopathy", 2: "moderate diabetic retinopathy",
3: "severe diabetic retinopathy", 4: "proliferative diabetic retinopathy"}
dataframe = pd.read_csv(PATH_DATASETS + path_dataset + "train.csv")
images_subpath = "train_images/"
data = []
for iFile in range(dataframe.shape[0]):
image_path = path_dataset + images_subpath + dataframe["id_code"][iFile] + ".png"
categories, atributes = [], []
categories.append(labels_dr[dataframe["diagnosis"][iFile]])
if os.path.isfile(PATH_DATASETS + image_path):
data.append({"image": image_path,
"atributes": atributes,
"categories": categories})
df_out = pd.DataFrame(data)
df_out.to_csv(PATH_DATAFRAME_PRETRAIN + "15_APTOS.csv")
def adequate_16_fundoct():
path_dataset = "16_FUND-OCT/"
data = []
dict_macula = {'acute CSR': 'acute central serous retinopathy', 'chronic CSR': 'chronic central serous retinopathy',
'ci-DME': 'cystoid macular edema', 'geographic_AMD': 'geographical age-related macular degeneration',
'Healthy': 'normal', 'neovascular_AMD': 'neovascular age-related macular degeneration'}
# Glaucoma/NoGlaucoma
subpath = "Dataset/OD/"
data = []
files = glob.glob(PATH_DATASETS + path_dataset + subpath + "*/*/*/*Color*")
for iFile in files:
data.append({"image": iFile.replace(PATH_DATASETS, ""),
"atributes": [],
"categories": [iFile.replace(PATH_DATASETS, "").split("/")[3].lower().replace("healthy", "normal")]})
# Macula-related
subpath = "Dataset/Macula/"
files = glob.glob(PATH_DATASETS + path_dataset + subpath + "*/*/*/*Color*")
for iFile in files:
data.append({"image": iFile.replace(PATH_DATASETS, ""),
"atributes": [],
"categories": [
dict_macula[iFile.replace(PATH_DATASETS, "").split("/")[3]]]})
df_out = pd.DataFrame(data)
df_out.to_csv(PATH_DATAFRAME_PRETRAIN + "16_FUND-OCT.csv")
def adequate_17_diaretdb1():
import xml.etree.ElementTree as ET
path_dataset = "17_DiaRetDB1/"
subpath_images = "documents/"
subpath_annotations = "groundtruth/"
files = os.listdir(PATH_DATASETS + path_dataset + subpath_images)
data = []
for iFile in files:
categories = []
for annotator in ["_01.xml","_02.xml","_03.xml","_04.xml"]:
annotation_id = iFile.replace(".png", annotator)
tree = ET.parse(PATH_DATASETS + path_dataset + subpath_annotations + annotation_id)
root = tree.getroot()
for item in root.findall('./markinglist/marking/'):
if item.tag == 'markingtype':
categories.append(item.text.lower().replace("_", " ").replace("irma", "intraretinal microvascular abnormalities"))
categories = list(np.unique(categories))
categories.remove("disc")
if len(categories) == 0:
continue
data.append({"image": path_dataset + subpath_images + iFile.replace(PATH_DATASETS, ""),
"atributes": [],
"categories": categories})
df_out = pd.DataFrame(data)
df_out.to_csv(PATH_DATAFRAME_PRETRAIN + "17_DiaRetDB1.csv")
def adequate_18_drions_db():
path_dataset = "18_DRIONS-DB/"
images_subpath = "documents/"
data = []
files = os.listdir(PATH_DATASETS + path_dataset + images_subpath)
for iFile in files:
data.append({"image": path_dataset + images_subpath + iFile,
"atributes": [],
"categories": ["no cataract", "a disease"]})
df_out = pd.DataFrame(data)
df_out.to_csv(PATH_DATAFRAME_PRETRAIN + "18_DRIONS-DB.csv")
def adequate_19_drishtigs1():
path_dataset = "19_Drishti-GS1/"
dataframe = pd.read_excel(PATH_DATASETS + path_dataset + "Drishti-GS1_diagnosis.xlsx", skiprows=3)[1:]
subpath_images = ["Drishti-GS1_files/Training/Images/", "Drishti-GS1_files/Test/Images/"]
data = []
for iPartition in subpath_images:
for iFile in range(dataframe.shape[0]):
id = dataframe["Drishti-GS File"].values[iFile][:-1]
finding = dataframe["Total"].values[iFile].lower()
image_path = path_dataset + iPartition + id + ".png"
if os.path.isfile(PATH_DATASETS + image_path):
data.append({"image": image_path,
"atributes": [],
"categories": [finding.replace("glaucomatous", "glaucoma")]})
df_out = pd.DataFrame(data)
df_out.to_csv(PATH_DATAFRAME_PRETRAIN + "19_Drishti-GS1.csv")
def adequate_20_e_ophta():
path_dataset = "20_E-ophta/"
labels = {"EX": "exudates", "healthy": "healthy", "MA": "microaneurysms"}
subpath_images = ["e_optha_EX/EX/", "e_optha_EX/healthy/", "e_optha_MA/MA/", "e_optha_EX/healthy/"]
data = []
for iSub in subpath_images:
finding = labels[iSub.split("/")[-2]].replace("healthy", "normal")
for root, dirs, files in os.walk(PATH_DATASETS + path_dataset + iSub):
for filename in files:
if filename != "Thumbs.db":
print(os.path.join(root, filename))
data.append({"image": os.path.join(root, filename).replace(PATH_DATASETS, ""),
"atributes": [],
"categories": [finding]})
df_out = pd.DataFrame(data)
df_out.to_csv(PATH_DATAFRAME_PRETRAIN + "20_E-ophta.csv")
def adequate_21_g1020():
path_dataset = "21_G1020/"
image_subpath = "Images/"
labels = {0: "normal", 1: "glaucoma"}
dataframe = pd.read_csv(PATH_DATASETS + path_dataset + "G1020.csv")
data = []
for iFile in range(dataframe.shape[0]):
id = dataframe["imageID"].values[iFile]
finding = labels[dataframe["binaryLabels"].values[iFile]]
image_path = path_dataset + image_subpath + id
if os.path.isfile(PATH_DATASETS + image_path):
data.append({"image": image_path,
"atributes": [],
"categories": [finding]})
df_out = pd.DataFrame(data)
df_out.to_csv(PATH_DATAFRAME_PRETRAIN + "21_G1020.csv")
def adequate_22_heimed():
path_dataset = "22_HEI-MED/"
data = []
return 1
def adequate_23_hrf():
path_dataset = "23_HRF/"
data = []
# Disease
image_subpath = "documents/"
labels = {"dr": "diabetic retinopathy", "g": "glaucoma", "h": "normal"}
files = os.listdir(PATH_DATASETS + path_dataset + image_subpath)
for iFile in files:
data.append({"image": path_dataset + image_subpath + iFile,
"atributes": [],
"categories": [labels[iFile.split("_")[-1].split(".")[0]]]})
# Noise
image_subpath = "Noise/"
labels = {"bad": "noisy", "good": "clean"}
files = os.listdir(PATH_DATASETS + path_dataset + image_subpath)
for iFile in files:
data.append({"image": path_dataset + image_subpath + iFile,
"atributes": [labels[iFile.split("_")[-1].split(".")[0]]],
"categories": []})
df_out = pd.DataFrame(data)
df_out.to_csv(PATH_DATAFRAME_PRETRAIN + "23_HRF.csv")
def adequate_24_origa():
path_dataset = "24_ORIGA/"
image_subpath = "Images/"
labels = {0: "no glaucoma", 1: "glaucoma"}
dataframe = pd.read_csv(PATH_DATASETS + path_dataset + "OrigaList.csv")
data = []
for iFile in range(dataframe.shape[0]):
id = dataframe["Filename"].values[iFile]
finding = labels[dataframe["Glaucoma"].values[iFile]]
image_path = path_dataset + image_subpath + id
if os.path.isfile(PATH_DATASETS + image_path):
data.append({"image": image_path,
"atributes": [],
"categories": [finding]})
df_out = pd.DataFrame(data)
df_out.to_csv(PATH_DATAFRAME_PRETRAIN + "24_ORIGA.csv")
def adequate_25_refuge():
path_dataset = "25_REFUGE/"
data = []
# Disease
image_subpath = "train/Images/" # We only have labels for train subset
labels = {"g": "glaucoma", "n": "no glaucoma"}
files = os.listdir(PATH_DATASETS + path_dataset + image_subpath)
for iFile in files:
data.append({"image": path_dataset + image_subpath + iFile,
"atributes": [],
"categories": [labels[iFile[0]]]})
df_out = pd.DataFrame(data)
df_out.to_csv(PATH_DATAFRAME_TRANSFERABILITY_CLASSIFICATION + "25_REFUGE.csv")
def adequate_26_roc():
path_dataset = "26_ROC/"
files = glob.glob(PATH_DATASETS + path_dataset + "*/*/*.jpg")
data = []
for iFile in files:
data.append({"image": iFile.replace(PATH_DATASETS,""),
"atributes": [],
"categories": ["microaneurysms"]})
df_out = pd.DataFrame(data)
df_out.to_csv(PATH_DATAFRAME_PRETRAIN + "26_ROC.csv")
def adequate_27_brset():
path_dataset = "27_BRSET/"
image_subpath = "fundus_photos/"
dataframe = pd.read_csv(PATH_DATASETS + path_dataset + "labels.csv")
anatomical_dict = {"1": "normal", "2": "abnormal", 'bv': ""}
dr_dict = {"0": "no diabetic retinopathy",
"1": "mild diabetic retinopathy",
"2": "moderate diabetic retinopathy",
"3": "severe diabetic retinopathy.",
"4": "proliferative diabetic retinopathy"}
findings = ["macular_edema", "scar", "nevus", "amd", "vascular_occlusion", "hypertensive_retinopathy",
"drusens", "hemorrhage", "retinal_detachment", "myopic_fundus", "increased_cup_disc", "other"]
find_names = ["macular edema", "scar", "nevus", "age-related macular degeneration", "vascular occlusion",
"hypertensive retinopathy", "drusens", "hemorrhage", "retina detachment",
"myopia", "increased cup disc", "a disease"]
data = []
for iFile in range(dataframe.shape[0]):
categories, atributes = [], []
id = dataframe["image_id"].values[iFile] + ".jpg"
# optic_disc
categories.append(anatomical_dict[dataframe["optic_disc"].values[iFile]] + " optic disc")
# vessels
categories.append(anatomical_dict[str(dataframe["vessels"].values[iFile])] + " vessels")
# macula
categories.append(anatomical_dict[str(dataframe["macula"].values[iFile])] + " macula")
# DR_ICDR
categories.append(dr_dict[str(dataframe["DR_ICDR"].values[iFile])])
# Noise
if dataframe["focus"].values[iFile] == 2 or dataframe["iluminaton"].values[iFile] == 2 \
or dataframe["image_field"].values[iFile] == 2 or dataframe["image_field"].values[iFile] == 2 :
atributes.append("Noisy")
# findings
for i in range(len(findings)):
if dataframe[findings[i]].values[iFile] == 1:
categories.append(find_names[i])
image_path = path_dataset + image_subpath + id
if os.path.isfile(PATH_DATASETS + image_path):
data.append({"image": image_path,
"atributes": [],
"categories": categories})
df_out = pd.DataFrame(data)
df_out.to_csv(PATH_DATAFRAME_PRETRAIN + "27_BRSET.csv")
def adequate_28_OIA():
path_dataset = "28_OIA-DDR/"
data = []
labels_dr = {0: "no diabetic retinopathy", 1: "mild diabetic retinopathy", 2: "moderate diabetic retinopathy",
3: "severe diabetic retinopathy", 4: "proliferative diabetic retinopathy", 5: ""}
subpath_grading = "DR_grading/"
subpath_segmentation = "lesion_segmentation/"
lesions_path = ["EX/", "HE/", "MA/", "SE/"]
lesions = ["hard exudates", "haemorrhages", "microaneurysms", "soft exudates"]
partitions = ["train", "test", "valid"]
for iPartition in partitions:
dataframe = pd.read_table(PATH_DATASETS + path_dataset + subpath_grading + iPartition + ".txt", delimiter=" ",
header=None)
files = list(dataframe[0].values)
for iFile in range(len(files)):
categories, atributes = [], []
image_path = path_dataset + subpath_grading + iPartition + "/" + files[iFile]
categories.append(labels_dr[dataframe[1].values[iFile]])
if dataframe[1].values[iFile] == 5:
atributes.append("noisy")
for i in range(len(lesions_path)):
for iiPartition in partitions:
if os.path.isfile(PATH_DATASETS + path_dataset + subpath_segmentation + iiPartition + "/" + "label/" + lesions_path[i] + files[iFile].replace(".jpg", ".tif")):
categories.append(lesions[i])
if os.path.isfile(PATH_DATASETS + image_path):
data.append({"image": image_path,
"atributes": atributes,
"categories": categories})
df_out = pd.DataFrame(data)
df_out.to_csv(PATH_DATAFRAME_PRETRAIN + "28_OIA-DDR.csv")
def adequate_29_airogs():
path_dataset = "29_AIROGS/"
image_subpath = "documents/" # We only have labels for train subset
labels = {"RG": "glaucoma", "NRG": "no glaucoma"}
dataframe = pd.read_csv(PATH_DATASETS + path_dataset + "train_labels.csv")
files = os.listdir(PATH_DATASETS + path_dataset + image_subpath)
data = []
for iFile in files[2:]:
print(iFile)
id = dataframe["challenge_id"] == iFile.split(".")[0]
finding = labels[dataframe[id]["class"].values[0]]
image_path = path_dataset + image_subpath + iFile
if os.path.isfile(PATH_DATASETS + image_path):
data.append({"image": image_path,
"atributes": [],
"categories": [finding]})
df_out = pd.DataFrame(data)
df_out.to_csv(PATH_DATAFRAME_PRETRAIN + "29_AIROGS.csv")
def adequate_30_sustech():
path_dataset = "30_SUSTech-SYSU/"
image_subpath = "originalImages/" # We only have labels for train subset
labels_dr = {0: "no diabetic retinopathy", 1: "mild diabetic retinopathy", 2: "moderate diabetic retinopathy",
3: "severe diabetic retinopathy", 4: "proliferative diabetic retinopathy", 5: ""}
dataframe = pd.read_csv(PATH_DATASETS + path_dataset + "Labels.csv")
files = os.listdir(PATH_DATASETS + path_dataset + image_subpath)
data = []
for iFile in files:
print(iFile)
id = dataframe["Fundus_images"] == iFile
if np.argwhere(id.values).__len__() > 0:
finding = labels_dr[dataframe[id]["DR_grade_American_Academy_of_Ophthalmology"].values[0]]
image_path = path_dataset + image_subpath + iFile
findings = [finding]
if os.path.isfile(PATH_DATASETS + image_path):
if os.path.isfile(PATH_DATASETS + path_dataset + "exudatesLabels/" + iFile.split(".")[0] + ".xml"):
findings.append("exudates")
data.append({"image": image_path,
"atributes": [],
"categories": findings})
df_out = pd.DataFrame(data)
df_out.to_csv(PATH_DATAFRAME_PRETRAIN + "30_SUSTech-SYSU.csv")
def adequate_31_jichi():
path_dataset = "31_JICHI/"
image_subpath = "documents/" # We only have labels for train subset
labels_dr = {"ndr": ["no diabetic retinopathy"],
"sdr": ["microaneurysm", "retinal hemorrhage", "hard exudate", "retinal edema",
"more than three small soft exudates"],
"ppdr": ["soft exudate", "varicose veins", "intraretinal microvascular abnormality",
"non-perfusion area over one disc area"],
"pdr": ["neovascularization", "preretinal haemorrhage", "fibrovascular proliferativemembrane",
"tractionalretinaldetachment"],
}
dataframe = pd.read_csv(PATH_DATASETS + path_dataset + "list.csv")
files = os.listdir(PATH_DATASETS + path_dataset + image_subpath)
data = []
for iFile in files:
print(iFile)
id = dataframe["Image"] == iFile
if np.argwhere(id.values).__len__() > 0:
finding = labels_dr[dataframe[id]["Davis_grading_of_one_figure"].values[0]]
image_path = path_dataset + image_subpath + iFile
if os.path.isfile(PATH_DATASETS + image_path):
data.append({"image": image_path,
"atributes": [],
"categories": finding})
df_out = pd.DataFrame(data)
df_out.to_csv(PATH_DATAFRAME_PRETRAIN + "31_JICHI.csv")
def adequate_32_chaksu():
path_dataset = "32_CHAKSU/"
subpaths = ["Train/", "Test/"]
data = []
scanners = ["Bosch", "Forus", "Remidio"]
dataframe_id = "Glaucoma_Decision_Comparison_[SCAN]_majority.csv"
path_images = "1.0_Original_Fundus_Images/"
labels = {"NORMAL": "no glaucoma", "GLAUCOMA SUSPECT": "glaucoma"}
formats = {'Bosch': '.JPG', 'Forus': '.png', 'Remidio': '.jpg'}
for iSubpath in subpaths:
for iScanner in scanners:
dataframe = pd.read_csv(PATH_DATASETS + path_dataset + iSubpath + dataframe_id.replace("[SCAN]", iScanner))
files = dataframe["Images"].values.tolist()
for iFile in files:
image_path = path_dataset + iSubpath + path_images + iScanner + "/" + iFile.split(".")[0] + formats[iScanner]