Image Segmentation
prasb commited on
Commit
d014c5a
·
verified ·
1 Parent(s): f61f78b

Upload 12 files

Browse files
.gitattributes CHANGED
@@ -33,3 +33,4 @@ saved_model/**/* filter=lfs diff=lfs merge=lfs -text
33
  *.zip filter=lfs diff=lfs merge=lfs -text
34
  *.zst filter=lfs diff=lfs merge=lfs -text
35
  *tfevents* filter=lfs diff=lfs merge=lfs -text
 
 
33
  *.zip filter=lfs diff=lfs merge=lfs -text
34
  *.zst filter=lfs diff=lfs merge=lfs -text
35
  *tfevents* filter=lfs diff=lfs merge=lfs -text
36
+ Dataset307_radiologist_annotations_Merlin_Plus/nnUNetTrainer__nnUNetPlannerResEncL_torchres_isotropic__3d_fullres/fold_all/progress.png filter=lfs diff=lfs merge=lfs -text
Dataset307_radiologist_annotations_Merlin_Plus/nnUNetTrainer__nnUNetPlannerResEncL_torchres_isotropic__3d_fullres/dataset.json ADDED
@@ -0,0 +1,89 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "channel_names": {
3
+ "0": "CT"
4
+ },
5
+ "labels": {
6
+ "background": 0,
7
+ "kidney_right": 1,
8
+ "kidney_left": 2,
9
+ "kidney_lesion": 3,
10
+ "kidney_lesion_kidney_right": 4,
11
+ "kidney_lesion_kidney_left": 5,
12
+ "pancreas": 6,
13
+ "pancreas_head": 7,
14
+ "pancreas_body": 8,
15
+ "pancreas_tail": 9,
16
+ "pancreatic_lesion": 10,
17
+ "pancreatic_lesion_pancreas_head": 11,
18
+ "pancreatic_lesion_pancreas_body": 12,
19
+ "pancreatic_lesion_pancreas_tail": 13,
20
+ "liver": 14,
21
+ "liver_segment_1": 15,
22
+ "liver_segment_2": 16,
23
+ "liver_segment_3": 17,
24
+ "liver_segment_4": 18,
25
+ "liver_segment_5": 19,
26
+ "liver_segment_6": 20,
27
+ "liver_segment_7": 21,
28
+ "liver_segment_8": 22,
29
+ "liver_lesion": 23,
30
+ "liver_lesion_liver_segment_1": 24,
31
+ "liver_lesion_liver_segment_2": 25,
32
+ "liver_lesion_liver_segment_3": 26,
33
+ "liver_lesion_liver_segment_4": 27,
34
+ "liver_lesion_liver_segment_5": 28,
35
+ "liver_lesion_liver_segment_6": 29,
36
+ "liver_lesion_liver_segment_7": 30,
37
+ "liver_lesion_liver_segment_8": 31,
38
+ "spleen": 32,
39
+ "spleen_lesion": 33,
40
+ "colon": 34,
41
+ "colon_lesion": 35,
42
+ "stomach": 36,
43
+ "stomach_lesion": 37,
44
+ "duodenum": 38,
45
+ "duodenum_lesion": 39,
46
+ "common_bile_duct": 40,
47
+ "intestine": 41,
48
+ "aorta": 42,
49
+ "postcava": 43,
50
+ "adrenal_gland_left": 44,
51
+ "adrenal_gland_right": 45,
52
+ "adrenal_lesion": 46,
53
+ "adrenal_lesion_adrenal_gland_left": 47,
54
+ "adrenal_lesion_adrenal_gland_right": 48,
55
+ "gall_bladder": 49,
56
+ "gallbladder_lesion": 50,
57
+ "gallbladder_lesion_gall_bladder": 51,
58
+ "bladder": 52,
59
+ "bladder_lesion": 53,
60
+ "celiac_trunk": 54,
61
+ "esophagus": 55,
62
+ "esophagus_lesion": 56,
63
+ "hepatic_vessel": 57,
64
+ "portal_vein_and_splenic_vein": 58,
65
+ "lung_left": 59,
66
+ "lung_right": 60,
67
+ "lung_lesion": 61,
68
+ "lung_lesion_lung_left": 62,
69
+ "lung_lesion_lung_right": 63,
70
+ "prostate": 64,
71
+ "prostate_lesion": 65,
72
+ "uterus_lesion": 66,
73
+ "uterus_lesion_prostate": 67,
74
+ "rectum": 68,
75
+ "femur_left": 69,
76
+ "femur_right": 70,
77
+ "superior_mesenteric_artery": 71,
78
+ "veins": 72,
79
+ "bone_lesion": 73,
80
+ "breast_lesion": 74
81
+ },
82
+ "numTraining": 423,
83
+ "file_ending": ".nii.gz",
84
+ "licence": "Whoever converted this dataset was lazy and didn't look it up!",
85
+ "converted_by": "Please enter your name, especially when sharing datasets with others in a common infrastructure!",
86
+ "citation": null,
87
+ "overwrite_image_reader_writer": "NibabelIOWithReorient",
88
+ "regions_class_order": "Dataset307_radiologist_annotations_Merlin_Plus"
89
+ }
Dataset307_radiologist_annotations_Merlin_Plus/nnUNetTrainer__nnUNetPlannerResEncL_torchres_isotropic__3d_fullres/dataset_fingerprint.json ADDED
@@ -0,0 +1,4248 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "foreground_intensity_properties_per_channel": {
3
+ "0": {
4
+ "max": 1000.0,
5
+ "mean": -128.6943359375,
6
+ "median": 51.01091003417969,
7
+ "min": -1000.0,
8
+ "percentile_00_5": -996.0021362304688,
9
+ "percentile_99_5": 851.0109252929688,
10
+ "std": 398.4421691894531
11
+ }
12
+ },
13
+ "median_relative_size_after_cropping": 1.0,
14
+ "shapes_after_crop": [
15
+ [
16
+ 332,
17
+ 512,
18
+ 512
19
+ ],
20
+ [
21
+ 409,
22
+ 512,
23
+ 512
24
+ ],
25
+ [
26
+ 416,
27
+ 512,
28
+ 512
29
+ ],
30
+ [
31
+ 343,
32
+ 512,
33
+ 512
34
+ ],
35
+ [
36
+ 310,
37
+ 512,
38
+ 512
39
+ ],
40
+ [
41
+ 451,
42
+ 512,
43
+ 512
44
+ ],
45
+ [
46
+ 379,
47
+ 512,
48
+ 512
49
+ ],
50
+ [
51
+ 357,
52
+ 512,
53
+ 512
54
+ ],
55
+ [
56
+ 456,
57
+ 512,
58
+ 512
59
+ ],
60
+ [
61
+ 446,
62
+ 512,
63
+ 512
64
+ ],
65
+ [
66
+ 468,
67
+ 512,
68
+ 512
69
+ ],
70
+ [
71
+ 417,
72
+ 512,
73
+ 512
74
+ ],
75
+ [
76
+ 474,
77
+ 512,
78
+ 512
79
+ ],
80
+ [
81
+ 310,
82
+ 512,
83
+ 512
84
+ ],
85
+ [
86
+ 473,
87
+ 512,
88
+ 512
89
+ ],
90
+ [
91
+ 442,
92
+ 512,
93
+ 512
94
+ ],
95
+ [
96
+ 394,
97
+ 512,
98
+ 512
99
+ ],
100
+ [
101
+ 375,
102
+ 512,
103
+ 512
104
+ ],
105
+ [
106
+ 355,
107
+ 512,
108
+ 512
109
+ ],
110
+ [
111
+ 653,
112
+ 512,
113
+ 512
114
+ ],
115
+ [
116
+ 445,
117
+ 512,
118
+ 512
119
+ ],
120
+ [
121
+ 472,
122
+ 512,
123
+ 512
124
+ ],
125
+ [
126
+ 327,
127
+ 512,
128
+ 512
129
+ ],
130
+ [
131
+ 479,
132
+ 512,
133
+ 512
134
+ ],
135
+ [
136
+ 492,
137
+ 512,
138
+ 512
139
+ ],
140
+ [
141
+ 356,
142
+ 512,
143
+ 512
144
+ ],
145
+ [
146
+ 492,
147
+ 512,
148
+ 512
149
+ ],
150
+ [
151
+ 455,
152
+ 512,
153
+ 512
154
+ ],
155
+ [
156
+ 590,
157
+ 512,
158
+ 512
159
+ ],
160
+ [
161
+ 518,
162
+ 512,
163
+ 512
164
+ ],
165
+ [
166
+ 395,
167
+ 512,
168
+ 512
169
+ ],
170
+ [
171
+ 507,
172
+ 512,
173
+ 512
174
+ ],
175
+ [
176
+ 345,
177
+ 512,
178
+ 512
179
+ ],
180
+ [
181
+ 557,
182
+ 512,
183
+ 512
184
+ ],
185
+ [
186
+ 434,
187
+ 512,
188
+ 512
189
+ ],
190
+ [
191
+ 464,
192
+ 512,
193
+ 512
194
+ ],
195
+ [
196
+ 449,
197
+ 512,
198
+ 512
199
+ ],
200
+ [
201
+ 422,
202
+ 512,
203
+ 512
204
+ ],
205
+ [
206
+ 480,
207
+ 512,
208
+ 512
209
+ ],
210
+ [
211
+ 493,
212
+ 512,
213
+ 512
214
+ ],
215
+ [
216
+ 343,
217
+ 512,
218
+ 512
219
+ ],
220
+ [
221
+ 378,
222
+ 512,
223
+ 512
224
+ ],
225
+ [
226
+ 452,
227
+ 512,
228
+ 512
229
+ ],
230
+ [
231
+ 445,
232
+ 512,
233
+ 512
234
+ ],
235
+ [
236
+ 370,
237
+ 512,
238
+ 512
239
+ ],
240
+ [
241
+ 354,
242
+ 512,
243
+ 512
244
+ ],
245
+ [
246
+ 370,
247
+ 512,
248
+ 512
249
+ ],
250
+ [
251
+ 499,
252
+ 512,
253
+ 512
254
+ ],
255
+ [
256
+ 411,
257
+ 512,
258
+ 512
259
+ ],
260
+ [
261
+ 484,
262
+ 512,
263
+ 512
264
+ ],
265
+ [
266
+ 367,
267
+ 512,
268
+ 512
269
+ ],
270
+ [
271
+ 467,
272
+ 512,
273
+ 512
274
+ ],
275
+ [
276
+ 348,
277
+ 512,
278
+ 512
279
+ ],
280
+ [
281
+ 465,
282
+ 512,
283
+ 512
284
+ ],
285
+ [
286
+ 476,
287
+ 512,
288
+ 512
289
+ ],
290
+ [
291
+ 385,
292
+ 512,
293
+ 512
294
+ ],
295
+ [
296
+ 397,
297
+ 512,
298
+ 512
299
+ ],
300
+ [
301
+ 511,
302
+ 512,
303
+ 512
304
+ ],
305
+ [
306
+ 503,
307
+ 512,
308
+ 512
309
+ ],
310
+ [
311
+ 420,
312
+ 512,
313
+ 512
314
+ ],
315
+ [
316
+ 389,
317
+ 512,
318
+ 512
319
+ ],
320
+ [
321
+ 501,
322
+ 512,
323
+ 512
324
+ ],
325
+ [
326
+ 485,
327
+ 512,
328
+ 512
329
+ ],
330
+ [
331
+ 425,
332
+ 512,
333
+ 512
334
+ ],
335
+ [
336
+ 469,
337
+ 512,
338
+ 512
339
+ ],
340
+ [
341
+ 422,
342
+ 512,
343
+ 512
344
+ ],
345
+ [
346
+ 373,
347
+ 512,
348
+ 512
349
+ ],
350
+ [
351
+ 477,
352
+ 512,
353
+ 512
354
+ ],
355
+ [
356
+ 382,
357
+ 512,
358
+ 512
359
+ ],
360
+ [
361
+ 418,
362
+ 512,
363
+ 512
364
+ ],
365
+ [
366
+ 447,
367
+ 512,
368
+ 512
369
+ ],
370
+ [
371
+ 429,
372
+ 512,
373
+ 512
374
+ ],
375
+ [
376
+ 353,
377
+ 512,
378
+ 512
379
+ ],
380
+ [
381
+ 301,
382
+ 512,
383
+ 512
384
+ ],
385
+ [
386
+ 281,
387
+ 512,
388
+ 512
389
+ ],
390
+ [
391
+ 355,
392
+ 512,
393
+ 512
394
+ ],
395
+ [
396
+ 466,
397
+ 512,
398
+ 512
399
+ ],
400
+ [
401
+ 385,
402
+ 512,
403
+ 512
404
+ ],
405
+ [
406
+ 369,
407
+ 512,
408
+ 512
409
+ ],
410
+ [
411
+ 442,
412
+ 512,
413
+ 512
414
+ ],
415
+ [
416
+ 464,
417
+ 512,
418
+ 512
419
+ ],
420
+ [
421
+ 392,
422
+ 512,
423
+ 512
424
+ ],
425
+ [
426
+ 502,
427
+ 512,
428
+ 512
429
+ ],
430
+ [
431
+ 335,
432
+ 512,
433
+ 512
434
+ ],
435
+ [
436
+ 452,
437
+ 512,
438
+ 512
439
+ ],
440
+ [
441
+ 462,
442
+ 512,
443
+ 512
444
+ ],
445
+ [
446
+ 383,
447
+ 512,
448
+ 512
449
+ ],
450
+ [
451
+ 458,
452
+ 512,
453
+ 512
454
+ ],
455
+ [
456
+ 387,
457
+ 512,
458
+ 512
459
+ ],
460
+ [
461
+ 386,
462
+ 512,
463
+ 512
464
+ ],
465
+ [
466
+ 349,
467
+ 512,
468
+ 512
469
+ ],
470
+ [
471
+ 549,
472
+ 512,
473
+ 512
474
+ ],
475
+ [
476
+ 339,
477
+ 512,
478
+ 512
479
+ ],
480
+ [
481
+ 489,
482
+ 512,
483
+ 512
484
+ ],
485
+ [
486
+ 410,
487
+ 512,
488
+ 512
489
+ ],
490
+ [
491
+ 444,
492
+ 512,
493
+ 512
494
+ ],
495
+ [
496
+ 441,
497
+ 512,
498
+ 512
499
+ ],
500
+ [
501
+ 417,
502
+ 512,
503
+ 512
504
+ ],
505
+ [
506
+ 344,
507
+ 512,
508
+ 512
509
+ ],
510
+ [
511
+ 417,
512
+ 512,
513
+ 512
514
+ ],
515
+ [
516
+ 388,
517
+ 512,
518
+ 512
519
+ ],
520
+ [
521
+ 381,
522
+ 512,
523
+ 512
524
+ ],
525
+ [
526
+ 377,
527
+ 512,
528
+ 512
529
+ ],
530
+ [
531
+ 393,
532
+ 512,
533
+ 512
534
+ ],
535
+ [
536
+ 346,
537
+ 512,
538
+ 512
539
+ ],
540
+ [
541
+ 452,
542
+ 512,
543
+ 512
544
+ ],
545
+ [
546
+ 428,
547
+ 512,
548
+ 512
549
+ ],
550
+ [
551
+ 493,
552
+ 512,
553
+ 512
554
+ ],
555
+ [
556
+ 409,
557
+ 512,
558
+ 512
559
+ ],
560
+ [
561
+ 501,
562
+ 512,
563
+ 512
564
+ ],
565
+ [
566
+ 355,
567
+ 512,
568
+ 512
569
+ ],
570
+ [
571
+ 429,
572
+ 512,
573
+ 512
574
+ ],
575
+ [
576
+ 373,
577
+ 512,
578
+ 512
579
+ ],
580
+ [
581
+ 329,
582
+ 512,
583
+ 512
584
+ ],
585
+ [
586
+ 316,
587
+ 512,
588
+ 512
589
+ ],
590
+ [
591
+ 470,
592
+ 512,
593
+ 512
594
+ ],
595
+ [
596
+ 425,
597
+ 512,
598
+ 512
599
+ ],
600
+ [
601
+ 389,
602
+ 512,
603
+ 512
604
+ ],
605
+ [
606
+ 388,
607
+ 512,
608
+ 512
609
+ ],
610
+ [
611
+ 315,
612
+ 512,
613
+ 512
614
+ ],
615
+ [
616
+ 450,
617
+ 512,
618
+ 512
619
+ ],
620
+ [
621
+ 455,
622
+ 512,
623
+ 512
624
+ ],
625
+ [
626
+ 421,
627
+ 512,
628
+ 512
629
+ ],
630
+ [
631
+ 365,
632
+ 512,
633
+ 512
634
+ ],
635
+ [
636
+ 386,
637
+ 512,
638
+ 512
639
+ ],
640
+ [
641
+ 405,
642
+ 512,
643
+ 512
644
+ ],
645
+ [
646
+ 396,
647
+ 512,
648
+ 512
649
+ ],
650
+ [
651
+ 446,
652
+ 512,
653
+ 512
654
+ ],
655
+ [
656
+ 169,
657
+ 512,
658
+ 512
659
+ ],
660
+ [
661
+ 466,
662
+ 512,
663
+ 512
664
+ ],
665
+ [
666
+ 518,
667
+ 512,
668
+ 512
669
+ ],
670
+ [
671
+ 510,
672
+ 512,
673
+ 512
674
+ ],
675
+ [
676
+ 482,
677
+ 512,
678
+ 512
679
+ ],
680
+ [
681
+ 376,
682
+ 512,
683
+ 512
684
+ ],
685
+ [
686
+ 465,
687
+ 512,
688
+ 512
689
+ ],
690
+ [
691
+ 360,
692
+ 512,
693
+ 512
694
+ ],
695
+ [
696
+ 403,
697
+ 512,
698
+ 512
699
+ ],
700
+ [
701
+ 532,
702
+ 512,
703
+ 512
704
+ ],
705
+ [
706
+ 405,
707
+ 512,
708
+ 512
709
+ ],
710
+ [
711
+ 369,
712
+ 512,
713
+ 512
714
+ ],
715
+ [
716
+ 429,
717
+ 512,
718
+ 512
719
+ ],
720
+ [
721
+ 409,
722
+ 512,
723
+ 512
724
+ ],
725
+ [
726
+ 452,
727
+ 512,
728
+ 512
729
+ ],
730
+ [
731
+ 378,
732
+ 512,
733
+ 512
734
+ ],
735
+ [
736
+ 381,
737
+ 512,
738
+ 512
739
+ ],
740
+ [
741
+ 365,
742
+ 512,
743
+ 512
744
+ ],
745
+ [
746
+ 353,
747
+ 512,
748
+ 512
749
+ ],
750
+ [
751
+ 513,
752
+ 512,
753
+ 512
754
+ ],
755
+ [
756
+ 345,
757
+ 512,
758
+ 512
759
+ ],
760
+ [
761
+ 394,
762
+ 512,
763
+ 512
764
+ ],
765
+ [
766
+ 494,
767
+ 512,
768
+ 512
769
+ ],
770
+ [
771
+ 409,
772
+ 512,
773
+ 512
774
+ ],
775
+ [
776
+ 377,
777
+ 512,
778
+ 512
779
+ ],
780
+ [
781
+ 386,
782
+ 512,
783
+ 512
784
+ ],
785
+ [
786
+ 445,
787
+ 512,
788
+ 512
789
+ ],
790
+ [
791
+ 391,
792
+ 512,
793
+ 512
794
+ ],
795
+ [
796
+ 381,
797
+ 512,
798
+ 512
799
+ ],
800
+ [
801
+ 464,
802
+ 512,
803
+ 512
804
+ ],
805
+ [
806
+ 399,
807
+ 512,
808
+ 512
809
+ ],
810
+ [
811
+ 382,
812
+ 512,
813
+ 512
814
+ ],
815
+ [
816
+ 521,
817
+ 512,
818
+ 512
819
+ ],
820
+ [
821
+ 502,
822
+ 512,
823
+ 512
824
+ ],
825
+ [
826
+ 352,
827
+ 512,
828
+ 512
829
+ ],
830
+ [
831
+ 517,
832
+ 512,
833
+ 512
834
+ ],
835
+ [
836
+ 392,
837
+ 512,
838
+ 512
839
+ ],
840
+ [
841
+ 365,
842
+ 512,
843
+ 512
844
+ ],
845
+ [
846
+ 409,
847
+ 512,
848
+ 512
849
+ ],
850
+ [
851
+ 462,
852
+ 512,
853
+ 512
854
+ ],
855
+ [
856
+ 411,
857
+ 512,
858
+ 512
859
+ ],
860
+ [
861
+ 348,
862
+ 512,
863
+ 512
864
+ ],
865
+ [
866
+ 461,
867
+ 512,
868
+ 512
869
+ ],
870
+ [
871
+ 345,
872
+ 512,
873
+ 512
874
+ ],
875
+ [
876
+ 362,
877
+ 512,
878
+ 512
879
+ ],
880
+ [
881
+ 410,
882
+ 512,
883
+ 512
884
+ ],
885
+ [
886
+ 511,
887
+ 512,
888
+ 512
889
+ ],
890
+ [
891
+ 352,
892
+ 512,
893
+ 512
894
+ ],
895
+ [
896
+ 341,
897
+ 512,
898
+ 512
899
+ ],
900
+ [
901
+ 380,
902
+ 512,
903
+ 512
904
+ ],
905
+ [
906
+ 411,
907
+ 512,
908
+ 512
909
+ ],
910
+ [
911
+ 496,
912
+ 512,
913
+ 512
914
+ ],
915
+ [
916
+ 371,
917
+ 512,
918
+ 512
919
+ ],
920
+ [
921
+ 422,
922
+ 512,
923
+ 512
924
+ ],
925
+ [
926
+ 407,
927
+ 512,
928
+ 512
929
+ ],
930
+ [
931
+ 383,
932
+ 512,
933
+ 512
934
+ ],
935
+ [
936
+ 386,
937
+ 512,
938
+ 512
939
+ ],
940
+ [
941
+ 467,
942
+ 512,
943
+ 512
944
+ ],
945
+ [
946
+ 401,
947
+ 512,
948
+ 512
949
+ ],
950
+ [
951
+ 443,
952
+ 512,
953
+ 512
954
+ ],
955
+ [
956
+ 349,
957
+ 512,
958
+ 512
959
+ ],
960
+ [
961
+ 445,
962
+ 512,
963
+ 512
964
+ ],
965
+ [
966
+ 334,
967
+ 512,
968
+ 512
969
+ ],
970
+ [
971
+ 412,
972
+ 512,
973
+ 512
974
+ ],
975
+ [
976
+ 339,
977
+ 512,
978
+ 512
979
+ ],
980
+ [
981
+ 421,
982
+ 512,
983
+ 512
984
+ ],
985
+ [
986
+ 413,
987
+ 512,
988
+ 512
989
+ ],
990
+ [
991
+ 390,
992
+ 512,
993
+ 512
994
+ ],
995
+ [
996
+ 396,
997
+ 512,
998
+ 512
999
+ ],
1000
+ [
1001
+ 409,
1002
+ 512,
1003
+ 512
1004
+ ],
1005
+ [
1006
+ 417,
1007
+ 512,
1008
+ 512
1009
+ ],
1010
+ [
1011
+ 383,
1012
+ 512,
1013
+ 512
1014
+ ],
1015
+ [
1016
+ 474,
1017
+ 512,
1018
+ 512
1019
+ ],
1020
+ [
1021
+ 371,
1022
+ 512,
1023
+ 512
1024
+ ],
1025
+ [
1026
+ 473,
1027
+ 512,
1028
+ 512
1029
+ ],
1030
+ [
1031
+ 370,
1032
+ 512,
1033
+ 512
1034
+ ],
1035
+ [
1036
+ 318,
1037
+ 512,
1038
+ 512
1039
+ ],
1040
+ [
1041
+ 453,
1042
+ 512,
1043
+ 512
1044
+ ],
1045
+ [
1046
+ 385,
1047
+ 512,
1048
+ 512
1049
+ ],
1050
+ [
1051
+ 389,
1052
+ 512,
1053
+ 512
1054
+ ],
1055
+ [
1056
+ 495,
1057
+ 512,
1058
+ 512
1059
+ ],
1060
+ [
1061
+ 353,
1062
+ 512,
1063
+ 512
1064
+ ],
1065
+ [
1066
+ 383,
1067
+ 512,
1068
+ 512
1069
+ ],
1070
+ [
1071
+ 436,
1072
+ 512,
1073
+ 512
1074
+ ],
1075
+ [
1076
+ 452,
1077
+ 512,
1078
+ 512
1079
+ ],
1080
+ [
1081
+ 451,
1082
+ 512,
1083
+ 512
1084
+ ],
1085
+ [
1086
+ 417,
1087
+ 512,
1088
+ 512
1089
+ ],
1090
+ [
1091
+ 388,
1092
+ 512,
1093
+ 512
1094
+ ],
1095
+ [
1096
+ 375,
1097
+ 512,
1098
+ 512
1099
+ ],
1100
+ [
1101
+ 426,
1102
+ 512,
1103
+ 512
1104
+ ],
1105
+ [
1106
+ 386,
1107
+ 512,
1108
+ 512
1109
+ ],
1110
+ [
1111
+ 393,
1112
+ 512,
1113
+ 512
1114
+ ],
1115
+ [
1116
+ 309,
1117
+ 512,
1118
+ 512
1119
+ ],
1120
+ [
1121
+ 409,
1122
+ 512,
1123
+ 512
1124
+ ],
1125
+ [
1126
+ 429,
1127
+ 512,
1128
+ 512
1129
+ ],
1130
+ [
1131
+ 365,
1132
+ 512,
1133
+ 512
1134
+ ],
1135
+ [
1136
+ 492,
1137
+ 512,
1138
+ 512
1139
+ ],
1140
+ [
1141
+ 394,
1142
+ 512,
1143
+ 512
1144
+ ],
1145
+ [
1146
+ 461,
1147
+ 512,
1148
+ 512
1149
+ ],
1150
+ [
1151
+ 539,
1152
+ 512,
1153
+ 512
1154
+ ],
1155
+ [
1156
+ 469,
1157
+ 512,
1158
+ 512
1159
+ ],
1160
+ [
1161
+ 471,
1162
+ 512,
1163
+ 512
1164
+ ],
1165
+ [
1166
+ 393,
1167
+ 512,
1168
+ 512
1169
+ ],
1170
+ [
1171
+ 456,
1172
+ 512,
1173
+ 512
1174
+ ],
1175
+ [
1176
+ 466,
1177
+ 512,
1178
+ 512
1179
+ ],
1180
+ [
1181
+ 349,
1182
+ 512,
1183
+ 512
1184
+ ],
1185
+ [
1186
+ 375,
1187
+ 512,
1188
+ 512
1189
+ ],
1190
+ [
1191
+ 446,
1192
+ 512,
1193
+ 512
1194
+ ],
1195
+ [
1196
+ 361,
1197
+ 512,
1198
+ 512
1199
+ ],
1200
+ [
1201
+ 401,
1202
+ 512,
1203
+ 512
1204
+ ],
1205
+ [
1206
+ 378,
1207
+ 512,
1208
+ 512
1209
+ ],
1210
+ [
1211
+ 413,
1212
+ 512,
1213
+ 512
1214
+ ],
1215
+ [
1216
+ 368,
1217
+ 512,
1218
+ 512
1219
+ ],
1220
+ [
1221
+ 504,
1222
+ 512,
1223
+ 512
1224
+ ],
1225
+ [
1226
+ 410,
1227
+ 512,
1228
+ 512
1229
+ ],
1230
+ [
1231
+ 447,
1232
+ 512,
1233
+ 512
1234
+ ],
1235
+ [
1236
+ 398,
1237
+ 512,
1238
+ 512
1239
+ ],
1240
+ [
1241
+ 363,
1242
+ 512,
1243
+ 512
1244
+ ],
1245
+ [
1246
+ 389,
1247
+ 512,
1248
+ 512
1249
+ ],
1250
+ [
1251
+ 472,
1252
+ 512,
1253
+ 512
1254
+ ],
1255
+ [
1256
+ 434,
1257
+ 512,
1258
+ 512
1259
+ ],
1260
+ [
1261
+ 385,
1262
+ 512,
1263
+ 512
1264
+ ],
1265
+ [
1266
+ 317,
1267
+ 512,
1268
+ 512
1269
+ ],
1270
+ [
1271
+ 454,
1272
+ 512,
1273
+ 512
1274
+ ],
1275
+ [
1276
+ 337,
1277
+ 512,
1278
+ 512
1279
+ ],
1280
+ [
1281
+ 381,
1282
+ 512,
1283
+ 512
1284
+ ],
1285
+ [
1286
+ 459,
1287
+ 512,
1288
+ 512
1289
+ ],
1290
+ [
1291
+ 393,
1292
+ 512,
1293
+ 512
1294
+ ],
1295
+ [
1296
+ 456,
1297
+ 512,
1298
+ 512
1299
+ ],
1300
+ [
1301
+ 425,
1302
+ 512,
1303
+ 512
1304
+ ],
1305
+ [
1306
+ 426,
1307
+ 512,
1308
+ 512
1309
+ ],
1310
+ [
1311
+ 409,
1312
+ 512,
1313
+ 512
1314
+ ],
1315
+ [
1316
+ 405,
1317
+ 512,
1318
+ 512
1319
+ ],
1320
+ [
1321
+ 332,
1322
+ 512,
1323
+ 512
1324
+ ],
1325
+ [
1326
+ 413,
1327
+ 512,
1328
+ 512
1329
+ ],
1330
+ [
1331
+ 492,
1332
+ 512,
1333
+ 512
1334
+ ],
1335
+ [
1336
+ 368,
1337
+ 512,
1338
+ 512
1339
+ ],
1340
+ [
1341
+ 325,
1342
+ 512,
1343
+ 512
1344
+ ],
1345
+ [
1346
+ 480,
1347
+ 512,
1348
+ 512
1349
+ ],
1350
+ [
1351
+ 482,
1352
+ 512,
1353
+ 512
1354
+ ],
1355
+ [
1356
+ 481,
1357
+ 512,
1358
+ 512
1359
+ ],
1360
+ [
1361
+ 380,
1362
+ 512,
1363
+ 512
1364
+ ],
1365
+ [
1366
+ 360,
1367
+ 512,
1368
+ 512
1369
+ ],
1370
+ [
1371
+ 422,
1372
+ 512,
1373
+ 512
1374
+ ],
1375
+ [
1376
+ 386,
1377
+ 512,
1378
+ 512
1379
+ ],
1380
+ [
1381
+ 358,
1382
+ 512,
1383
+ 512
1384
+ ],
1385
+ [
1386
+ 497,
1387
+ 512,
1388
+ 512
1389
+ ],
1390
+ [
1391
+ 437,
1392
+ 512,
1393
+ 512
1394
+ ],
1395
+ [
1396
+ 355,
1397
+ 512,
1398
+ 512
1399
+ ],
1400
+ [
1401
+ 349,
1402
+ 512,
1403
+ 512
1404
+ ],
1405
+ [
1406
+ 390,
1407
+ 512,
1408
+ 512
1409
+ ],
1410
+ [
1411
+ 497,
1412
+ 512,
1413
+ 512
1414
+ ],
1415
+ [
1416
+ 452,
1417
+ 512,
1418
+ 512
1419
+ ],
1420
+ [
1421
+ 403,
1422
+ 512,
1423
+ 512
1424
+ ],
1425
+ [
1426
+ 409,
1427
+ 512,
1428
+ 512
1429
+ ],
1430
+ [
1431
+ 533,
1432
+ 512,
1433
+ 512
1434
+ ],
1435
+ [
1436
+ 351,
1437
+ 512,
1438
+ 512
1439
+ ],
1440
+ [
1441
+ 337,
1442
+ 512,
1443
+ 512
1444
+ ],
1445
+ [
1446
+ 411,
1447
+ 512,
1448
+ 512
1449
+ ],
1450
+ [
1451
+ 358,
1452
+ 512,
1453
+ 512
1454
+ ],
1455
+ [
1456
+ 398,
1457
+ 512,
1458
+ 512
1459
+ ],
1460
+ [
1461
+ 377,
1462
+ 512,
1463
+ 512
1464
+ ],
1465
+ [
1466
+ 507,
1467
+ 512,
1468
+ 512
1469
+ ],
1470
+ [
1471
+ 415,
1472
+ 512,
1473
+ 512
1474
+ ],
1475
+ [
1476
+ 357,
1477
+ 512,
1478
+ 512
1479
+ ],
1480
+ [
1481
+ 373,
1482
+ 512,
1483
+ 512
1484
+ ],
1485
+ [
1486
+ 387,
1487
+ 512,
1488
+ 512
1489
+ ],
1490
+ [
1491
+ 435,
1492
+ 512,
1493
+ 512
1494
+ ],
1495
+ [
1496
+ 106,
1497
+ 512,
1498
+ 512
1499
+ ],
1500
+ [
1501
+ 371,
1502
+ 512,
1503
+ 512
1504
+ ],
1505
+ [
1506
+ 464,
1507
+ 512,
1508
+ 512
1509
+ ],
1510
+ [
1511
+ 381,
1512
+ 512,
1513
+ 512
1514
+ ],
1515
+ [
1516
+ 538,
1517
+ 512,
1518
+ 512
1519
+ ],
1520
+ [
1521
+ 367,
1522
+ 512,
1523
+ 512
1524
+ ],
1525
+ [
1526
+ 386,
1527
+ 512,
1528
+ 512
1529
+ ],
1530
+ [
1531
+ 329,
1532
+ 512,
1533
+ 512
1534
+ ],
1535
+ [
1536
+ 408,
1537
+ 512,
1538
+ 512
1539
+ ],
1540
+ [
1541
+ 321,
1542
+ 512,
1543
+ 512
1544
+ ],
1545
+ [
1546
+ 511,
1547
+ 512,
1548
+ 512
1549
+ ],
1550
+ [
1551
+ 374,
1552
+ 512,
1553
+ 512
1554
+ ],
1555
+ [
1556
+ 436,
1557
+ 512,
1558
+ 512
1559
+ ],
1560
+ [
1561
+ 524,
1562
+ 512,
1563
+ 512
1564
+ ],
1565
+ [
1566
+ 476,
1567
+ 512,
1568
+ 512
1569
+ ],
1570
+ [
1571
+ 400,
1572
+ 512,
1573
+ 512
1574
+ ],
1575
+ [
1576
+ 83,
1577
+ 512,
1578
+ 512
1579
+ ],
1580
+ [
1581
+ 330,
1582
+ 512,
1583
+ 512
1584
+ ],
1585
+ [
1586
+ 345,
1587
+ 512,
1588
+ 512
1589
+ ],
1590
+ [
1591
+ 357,
1592
+ 512,
1593
+ 512
1594
+ ],
1595
+ [
1596
+ 353,
1597
+ 512,
1598
+ 512
1599
+ ],
1600
+ [
1601
+ 397,
1602
+ 512,
1603
+ 512
1604
+ ],
1605
+ [
1606
+ 429,
1607
+ 512,
1608
+ 512
1609
+ ],
1610
+ [
1611
+ 466,
1612
+ 512,
1613
+ 512
1614
+ ],
1615
+ [
1616
+ 486,
1617
+ 512,
1618
+ 512
1619
+ ],
1620
+ [
1621
+ 297,
1622
+ 512,
1623
+ 512
1624
+ ],
1625
+ [
1626
+ 448,
1627
+ 512,
1628
+ 512
1629
+ ],
1630
+ [
1631
+ 480,
1632
+ 512,
1633
+ 512
1634
+ ],
1635
+ [
1636
+ 427,
1637
+ 512,
1638
+ 512
1639
+ ],
1640
+ [
1641
+ 361,
1642
+ 512,
1643
+ 512
1644
+ ],
1645
+ [
1646
+ 322,
1647
+ 512,
1648
+ 512
1649
+ ],
1650
+ [
1651
+ 421,
1652
+ 512,
1653
+ 512
1654
+ ],
1655
+ [
1656
+ 411,
1657
+ 512,
1658
+ 512
1659
+ ],
1660
+ [
1661
+ 427,
1662
+ 512,
1663
+ 512
1664
+ ],
1665
+ [
1666
+ 409,
1667
+ 512,
1668
+ 512
1669
+ ],
1670
+ [
1671
+ 438,
1672
+ 512,
1673
+ 512
1674
+ ],
1675
+ [
1676
+ 442,
1677
+ 512,
1678
+ 512
1679
+ ],
1680
+ [
1681
+ 344,
1682
+ 512,
1683
+ 512
1684
+ ],
1685
+ [
1686
+ 426,
1687
+ 512,
1688
+ 512
1689
+ ],
1690
+ [
1691
+ 384,
1692
+ 512,
1693
+ 512
1694
+ ],
1695
+ [
1696
+ 436,
1697
+ 512,
1698
+ 512
1699
+ ],
1700
+ [
1701
+ 369,
1702
+ 512,
1703
+ 512
1704
+ ],
1705
+ [
1706
+ 465,
1707
+ 512,
1708
+ 512
1709
+ ],
1710
+ [
1711
+ 409,
1712
+ 512,
1713
+ 512
1714
+ ],
1715
+ [
1716
+ 404,
1717
+ 512,
1718
+ 512
1719
+ ],
1720
+ [
1721
+ 385,
1722
+ 512,
1723
+ 512
1724
+ ],
1725
+ [
1726
+ 337,
1727
+ 512,
1728
+ 512
1729
+ ],
1730
+ [
1731
+ 421,
1732
+ 512,
1733
+ 512
1734
+ ],
1735
+ [
1736
+ 486,
1737
+ 512,
1738
+ 512
1739
+ ],
1740
+ [
1741
+ 409,
1742
+ 512,
1743
+ 512
1744
+ ],
1745
+ [
1746
+ 464,
1747
+ 512,
1748
+ 512
1749
+ ],
1750
+ [
1751
+ 378,
1752
+ 512,
1753
+ 512
1754
+ ],
1755
+ [
1756
+ 391,
1757
+ 512,
1758
+ 512
1759
+ ],
1760
+ [
1761
+ 391,
1762
+ 512,
1763
+ 512
1764
+ ],
1765
+ [
1766
+ 494,
1767
+ 512,
1768
+ 512
1769
+ ],
1770
+ [
1771
+ 391,
1772
+ 512,
1773
+ 512
1774
+ ],
1775
+ [
1776
+ 496,
1777
+ 512,
1778
+ 512
1779
+ ],
1780
+ [
1781
+ 504,
1782
+ 512,
1783
+ 512
1784
+ ],
1785
+ [
1786
+ 363,
1787
+ 512,
1788
+ 512
1789
+ ],
1790
+ [
1791
+ 349,
1792
+ 512,
1793
+ 512
1794
+ ],
1795
+ [
1796
+ 344,
1797
+ 512,
1798
+ 512
1799
+ ],
1800
+ [
1801
+ 403,
1802
+ 512,
1803
+ 512
1804
+ ],
1805
+ [
1806
+ 424,
1807
+ 512,
1808
+ 512
1809
+ ],
1810
+ [
1811
+ 393,
1812
+ 512,
1813
+ 512
1814
+ ],
1815
+ [
1816
+ 401,
1817
+ 512,
1818
+ 512
1819
+ ],
1820
+ [
1821
+ 421,
1822
+ 512,
1823
+ 512
1824
+ ],
1825
+ [
1826
+ 376,
1827
+ 512,
1828
+ 512
1829
+ ],
1830
+ [
1831
+ 357,
1832
+ 512,
1833
+ 512
1834
+ ],
1835
+ [
1836
+ 463,
1837
+ 512,
1838
+ 512
1839
+ ],
1840
+ [
1841
+ 424,
1842
+ 512,
1843
+ 512
1844
+ ],
1845
+ [
1846
+ 388,
1847
+ 512,
1848
+ 512
1849
+ ],
1850
+ [
1851
+ 473,
1852
+ 512,
1853
+ 512
1854
+ ],
1855
+ [
1856
+ 427,
1857
+ 512,
1858
+ 512
1859
+ ],
1860
+ [
1861
+ 449,
1862
+ 512,
1863
+ 512
1864
+ ],
1865
+ [
1866
+ 480,
1867
+ 512,
1868
+ 512
1869
+ ],
1870
+ [
1871
+ 458,
1872
+ 512,
1873
+ 512
1874
+ ],
1875
+ [
1876
+ 527,
1877
+ 512,
1878
+ 512
1879
+ ],
1880
+ [
1881
+ 392,
1882
+ 512,
1883
+ 512
1884
+ ],
1885
+ [
1886
+ 432,
1887
+ 512,
1888
+ 512
1889
+ ],
1890
+ [
1891
+ 945,
1892
+ 512,
1893
+ 512
1894
+ ],
1895
+ [
1896
+ 369,
1897
+ 512,
1898
+ 512
1899
+ ],
1900
+ [
1901
+ 474,
1902
+ 512,
1903
+ 512
1904
+ ],
1905
+ [
1906
+ 385,
1907
+ 512,
1908
+ 512
1909
+ ],
1910
+ [
1911
+ 333,
1912
+ 512,
1913
+ 512
1914
+ ],
1915
+ [
1916
+ 539,
1917
+ 512,
1918
+ 512
1919
+ ],
1920
+ [
1921
+ 417,
1922
+ 512,
1923
+ 512
1924
+ ],
1925
+ [
1926
+ 395,
1927
+ 512,
1928
+ 512
1929
+ ],
1930
+ [
1931
+ 417,
1932
+ 512,
1933
+ 512
1934
+ ],
1935
+ [
1936
+ 321,
1937
+ 512,
1938
+ 512
1939
+ ],
1940
+ [
1941
+ 496,
1942
+ 512,
1943
+ 512
1944
+ ],
1945
+ [
1946
+ 492,
1947
+ 512,
1948
+ 512
1949
+ ],
1950
+ [
1951
+ 411,
1952
+ 512,
1953
+ 512
1954
+ ],
1955
+ [
1956
+ 331,
1957
+ 512,
1958
+ 512
1959
+ ],
1960
+ [
1961
+ 373,
1962
+ 512,
1963
+ 512
1964
+ ],
1965
+ [
1966
+ 410,
1967
+ 512,
1968
+ 512
1969
+ ],
1970
+ [
1971
+ 455,
1972
+ 512,
1973
+ 512
1974
+ ],
1975
+ [
1976
+ 410,
1977
+ 512,
1978
+ 512
1979
+ ],
1980
+ [
1981
+ 537,
1982
+ 512,
1983
+ 512
1984
+ ],
1985
+ [
1986
+ 389,
1987
+ 512,
1988
+ 512
1989
+ ],
1990
+ [
1991
+ 344,
1992
+ 512,
1993
+ 512
1994
+ ],
1995
+ [
1996
+ 516,
1997
+ 512,
1998
+ 512
1999
+ ],
2000
+ [
2001
+ 405,
2002
+ 512,
2003
+ 512
2004
+ ],
2005
+ [
2006
+ 415,
2007
+ 512,
2008
+ 512
2009
+ ],
2010
+ [
2011
+ 452,
2012
+ 512,
2013
+ 512
2014
+ ],
2015
+ [
2016
+ 409,
2017
+ 512,
2018
+ 512
2019
+ ],
2020
+ [
2021
+ 444,
2022
+ 512,
2023
+ 512
2024
+ ],
2025
+ [
2026
+ 385,
2027
+ 512,
2028
+ 512
2029
+ ],
2030
+ [
2031
+ 445,
2032
+ 512,
2033
+ 512
2034
+ ],
2035
+ [
2036
+ 389,
2037
+ 512,
2038
+ 512
2039
+ ],
2040
+ [
2041
+ 365,
2042
+ 512,
2043
+ 512
2044
+ ],
2045
+ [
2046
+ 386,
2047
+ 512,
2048
+ 512
2049
+ ],
2050
+ [
2051
+ 417,
2052
+ 512,
2053
+ 512
2054
+ ],
2055
+ [
2056
+ 365,
2057
+ 512,
2058
+ 512
2059
+ ],
2060
+ [
2061
+ 381,
2062
+ 512,
2063
+ 512
2064
+ ],
2065
+ [
2066
+ 457,
2067
+ 512,
2068
+ 512
2069
+ ],
2070
+ [
2071
+ 365,
2072
+ 512,
2073
+ 512
2074
+ ],
2075
+ [
2076
+ 403,
2077
+ 512,
2078
+ 512
2079
+ ],
2080
+ [
2081
+ 504,
2082
+ 512,
2083
+ 512
2084
+ ],
2085
+ [
2086
+ 405,
2087
+ 512,
2088
+ 512
2089
+ ],
2090
+ [
2091
+ 491,
2092
+ 512,
2093
+ 512
2094
+ ],
2095
+ [
2096
+ 480,
2097
+ 512,
2098
+ 512
2099
+ ],
2100
+ [
2101
+ 351,
2102
+ 512,
2103
+ 512
2104
+ ],
2105
+ [
2106
+ 441,
2107
+ 512,
2108
+ 512
2109
+ ],
2110
+ [
2111
+ 478,
2112
+ 512,
2113
+ 512
2114
+ ],
2115
+ [
2116
+ 384,
2117
+ 512,
2118
+ 512
2119
+ ],
2120
+ [
2121
+ 397,
2122
+ 512,
2123
+ 512
2124
+ ],
2125
+ [
2126
+ 394,
2127
+ 512,
2128
+ 512
2129
+ ]
2130
+ ],
2131
+ "spacings": [
2132
+ [
2133
+ 1.25,
2134
+ 0.7910159826278687,
2135
+ 0.7910159826278687
2136
+ ],
2137
+ [
2138
+ 1.25,
2139
+ 0.7363280057907104,
2140
+ 0.7363280057907104
2141
+ ],
2142
+ [
2143
+ 1.0,
2144
+ 0.796875,
2145
+ 0.796875
2146
+ ],
2147
+ [
2148
+ 1.25,
2149
+ 0.625,
2150
+ 0.625
2151
+ ],
2152
+ [
2153
+ 1.25,
2154
+ 0.6035159826278687,
2155
+ 0.6035159826278687
2156
+ ],
2157
+ [
2158
+ 1.0,
2159
+ 0.927734375,
2160
+ 0.927734375
2161
+ ],
2162
+ [
2163
+ 1.25,
2164
+ 0.8691409826278687,
2165
+ 0.8691409826278687
2166
+ ],
2167
+ [
2168
+ 1.25,
2169
+ 0.625,
2170
+ 0.625
2171
+ ],
2172
+ [
2173
+ 1.0,
2174
+ 0.763671875,
2175
+ 0.763671875
2176
+ ],
2177
+ [
2178
+ 1.0,
2179
+ 0.78125,
2180
+ 0.78125
2181
+ ],
2182
+ [
2183
+ 1.0,
2184
+ 0.66796875,
2185
+ 0.66796875
2186
+ ],
2187
+ [
2188
+ 1.0,
2189
+ 0.712890625,
2190
+ 0.712890625
2191
+ ],
2192
+ [
2193
+ 1.0,
2194
+ 0.802734375,
2195
+ 0.802734375
2196
+ ],
2197
+ [
2198
+ 1.25,
2199
+ 0.6875,
2200
+ 0.6875
2201
+ ],
2202
+ [
2203
+ 1.0,
2204
+ 0.7421875,
2205
+ 0.7421875
2206
+ ],
2207
+ [
2208
+ 1.0,
2209
+ 0.6953125,
2210
+ 0.6953125
2211
+ ],
2212
+ [
2213
+ 1.25,
2214
+ 0.9765620231628418,
2215
+ 0.9765620231628418
2216
+ ],
2217
+ [
2218
+ 1.25,
2219
+ 0.703125,
2220
+ 0.703125
2221
+ ],
2222
+ [
2223
+ 1.25,
2224
+ 0.6835939884185791,
2225
+ 0.6835939884185791
2226
+ ],
2227
+ [
2228
+ 0.625,
2229
+ 0.703125,
2230
+ 0.703125
2231
+ ],
2232
+ [
2233
+ 1.25,
2234
+ 0.9628909826278687,
2235
+ 0.9628909826278687
2236
+ ],
2237
+ [
2238
+ 1.0,
2239
+ 0.6953125,
2240
+ 0.6953125
2241
+ ],
2242
+ [
2243
+ 1.25,
2244
+ 0.7851560115814209,
2245
+ 0.7851560115814209
2246
+ ],
2247
+ [
2248
+ 1.0,
2249
+ 0.75390625,
2250
+ 0.75390625
2251
+ ],
2252
+ [
2253
+ 1.0,
2254
+ 0.755859375,
2255
+ 0.755859375
2256
+ ],
2257
+ [
2258
+ 1.25,
2259
+ 0.703125,
2260
+ 0.703125
2261
+ ],
2262
+ [
2263
+ 1.0,
2264
+ 0.771484375,
2265
+ 0.771484375
2266
+ ],
2267
+ [
2268
+ 1.0,
2269
+ 0.8046875,
2270
+ 0.8046875
2271
+ ],
2272
+ [
2273
+ 1.0,
2274
+ 0.90234375,
2275
+ 0.90234375
2276
+ ],
2277
+ [
2278
+ 1.0,
2279
+ 0.75390625,
2280
+ 0.75390625
2281
+ ],
2282
+ [
2283
+ 1.25,
2284
+ 0.8496090173721313,
2285
+ 0.8496090173721313
2286
+ ],
2287
+ [
2288
+ 1.0,
2289
+ 0.7421875,
2290
+ 0.7421875
2291
+ ],
2292
+ [
2293
+ 1.25,
2294
+ 0.703125,
2295
+ 0.703125
2296
+ ],
2297
+ [
2298
+ 0.800000011920929,
2299
+ 0.78125,
2300
+ 0.78125
2301
+ ],
2302
+ [
2303
+ 1.0,
2304
+ 0.646484375,
2305
+ 0.646484375
2306
+ ],
2307
+ [
2308
+ 1.0,
2309
+ 0.677734375,
2310
+ 0.677734375
2311
+ ],
2312
+ [
2313
+ 1.0,
2314
+ 0.693359375,
2315
+ 0.693359375
2316
+ ],
2317
+ [
2318
+ 1.0,
2319
+ 0.580078125,
2320
+ 0.580078125
2321
+ ],
2322
+ [
2323
+ 1.0,
2324
+ 0.79296875,
2325
+ 0.79296875
2326
+ ],
2327
+ [
2328
+ 1.0,
2329
+ 0.69140625,
2330
+ 0.69140625
2331
+ ],
2332
+ [
2333
+ 1.25,
2334
+ 0.9765620231628418,
2335
+ 0.9765620231628418
2336
+ ],
2337
+ [
2338
+ 1.25,
2339
+ 0.703125,
2340
+ 0.703125
2341
+ ],
2342
+ [
2343
+ 1.0,
2344
+ 0.74609375,
2345
+ 0.74609375
2346
+ ],
2347
+ [
2348
+ 1.25,
2349
+ 0.796875,
2350
+ 0.796875
2351
+ ],
2352
+ [
2353
+ 1.0,
2354
+ 0.654296875,
2355
+ 0.654296875
2356
+ ],
2357
+ [
2358
+ 1.25,
2359
+ 0.71875,
2360
+ 0.71875
2361
+ ],
2362
+ [
2363
+ 1.25,
2364
+ 0.7246090173721313,
2365
+ 0.7246090173721313
2366
+ ],
2367
+ [
2368
+ 1.0,
2369
+ 0.794921875,
2370
+ 0.794921875
2371
+ ],
2372
+ [
2373
+ 1.0,
2374
+ 0.732421875,
2375
+ 0.732421875
2376
+ ],
2377
+ [
2378
+ 1.0,
2379
+ 0.67578125,
2380
+ 0.67578125
2381
+ ],
2382
+ [
2383
+ 1.25,
2384
+ 0.7851560115814209,
2385
+ 0.7851560115814209
2386
+ ],
2387
+ [
2388
+ 1.0,
2389
+ 0.716796875,
2390
+ 0.716796875
2391
+ ],
2392
+ [
2393
+ 1.25,
2394
+ 0.7578120231628418,
2395
+ 0.7578120231628418
2396
+ ],
2397
+ [
2398
+ 1.0,
2399
+ 0.662109375,
2400
+ 0.662109375
2401
+ ],
2402
+ [
2403
+ 1.0,
2404
+ 0.666015625,
2405
+ 0.666015625
2406
+ ],
2407
+ [
2408
+ 1.25,
2409
+ 0.7617189884185791,
2410
+ 0.7617189884185791
2411
+ ],
2412
+ [
2413
+ 1.25,
2414
+ 0.75,
2415
+ 0.75
2416
+ ],
2417
+ [
2418
+ 1.0,
2419
+ 0.755859375,
2420
+ 0.755859375
2421
+ ],
2422
+ [
2423
+ 1.0,
2424
+ 0.771484375,
2425
+ 0.771484375
2426
+ ],
2427
+ [
2428
+ 1.0,
2429
+ 0.73828125,
2430
+ 0.73828125
2431
+ ],
2432
+ [
2433
+ 1.25,
2434
+ 0.6386719942092896,
2435
+ 0.6386719942092896
2436
+ ],
2437
+ [
2438
+ 1.0,
2439
+ 0.83203125,
2440
+ 0.83203125
2441
+ ],
2442
+ [
2443
+ 1.0,
2444
+ 0.7265625,
2445
+ 0.7265625
2446
+ ],
2447
+ [
2448
+ 1.25,
2449
+ 0.9765620231628418,
2450
+ 0.9765620231628418
2451
+ ],
2452
+ [
2453
+ 1.0,
2454
+ 0.640625,
2455
+ 0.640625
2456
+ ],
2457
+ [
2458
+ 1.0,
2459
+ 0.720703125,
2460
+ 0.720703125
2461
+ ],
2462
+ [
2463
+ 1.25,
2464
+ 0.8085939884185791,
2465
+ 0.8085939884185791
2466
+ ],
2467
+ [
2468
+ 1.0,
2469
+ 0.6484375,
2470
+ 0.6484375
2471
+ ],
2472
+ [
2473
+ 1.25,
2474
+ 0.7421879768371582,
2475
+ 0.7421879768371582
2476
+ ],
2477
+ [
2478
+ 1.25,
2479
+ 0.8652340173721313,
2480
+ 0.8652340173721313
2481
+ ],
2482
+ [
2483
+ 1.0,
2484
+ 0.6484375,
2485
+ 0.6484375
2486
+ ],
2487
+ [
2488
+ 1.25,
2489
+ 0.8164060115814209,
2490
+ 0.8164060115814209
2491
+ ],
2492
+ [
2493
+ 1.25,
2494
+ 0.75,
2495
+ 0.75
2496
+ ],
2497
+ [
2498
+ 1.25,
2499
+ 0.7226560115814209,
2500
+ 0.7226560115814209
2501
+ ],
2502
+ [
2503
+ 1.25,
2504
+ 0.703125,
2505
+ 0.703125
2506
+ ],
2507
+ [
2508
+ 1.25,
2509
+ 0.6230469942092896,
2510
+ 0.6230469942092896
2511
+ ],
2512
+ [
2513
+ 1.0,
2514
+ 0.640625,
2515
+ 0.640625
2516
+ ],
2517
+ [
2518
+ 1.25,
2519
+ 0.703125,
2520
+ 0.703125
2521
+ ],
2522
+ [
2523
+ 1.25,
2524
+ 0.703125,
2525
+ 0.703125
2526
+ ],
2527
+ [
2528
+ 1.0,
2529
+ 0.8671875,
2530
+ 0.8671875
2531
+ ],
2532
+ [
2533
+ 1.0,
2534
+ 0.7734375,
2535
+ 0.7734375
2536
+ ],
2537
+ [
2538
+ 1.25,
2539
+ 0.7246090173721313,
2540
+ 0.7246090173721313
2541
+ ],
2542
+ [
2543
+ 1.0,
2544
+ 0.791015625,
2545
+ 0.791015625
2546
+ ],
2547
+ [
2548
+ 1.25,
2549
+ 0.703125,
2550
+ 0.703125
2551
+ ],
2552
+ [
2553
+ 1.0,
2554
+ 0.541015625,
2555
+ 0.541015625
2556
+ ],
2557
+ [
2558
+ 1.0,
2559
+ 0.765625,
2560
+ 0.765625
2561
+ ],
2562
+ [
2563
+ 1.25,
2564
+ 0.9257810115814209,
2565
+ 0.9257810115814209
2566
+ ],
2567
+ [
2568
+ 1.0,
2569
+ 0.64453125,
2570
+ 0.64453125
2571
+ ],
2572
+ [
2573
+ 1.25,
2574
+ 0.703125,
2575
+ 0.703125
2576
+ ],
2577
+ [
2578
+ 1.25,
2579
+ 0.703125,
2580
+ 0.703125
2581
+ ],
2582
+ [
2583
+ 1.25,
2584
+ 0.6328120231628418,
2585
+ 0.6328120231628418
2586
+ ],
2587
+ [
2588
+ 0.800000011920929,
2589
+ 0.6835939884185791,
2590
+ 0.6835939884185791
2591
+ ],
2592
+ [
2593
+ 1.25,
2594
+ 0.765625,
2595
+ 0.765625
2596
+ ],
2597
+ [
2598
+ 1.0,
2599
+ 0.875,
2600
+ 0.875
2601
+ ],
2602
+ [
2603
+ 1.25,
2604
+ 0.9472659826278687,
2605
+ 0.9472659826278687
2606
+ ],
2607
+ [
2608
+ 1.25,
2609
+ 0.7988280057907104,
2610
+ 0.7988280057907104
2611
+ ],
2612
+ [
2613
+ 1.0,
2614
+ 0.73828125,
2615
+ 0.73828125
2616
+ ],
2617
+ [
2618
+ 1.25,
2619
+ 0.8457030057907104,
2620
+ 0.8457030057907104
2621
+ ],
2622
+ [
2623
+ 1.25,
2624
+ 0.8652340173721313,
2625
+ 0.8652340173721313
2626
+ ],
2627
+ [
2628
+ 1.25,
2629
+ 0.9277340173721313,
2630
+ 0.9277340173721313
2631
+ ],
2632
+ [
2633
+ 1.25,
2634
+ 0.8125,
2635
+ 0.8125
2636
+ ],
2637
+ [
2638
+ 1.25,
2639
+ 0.6777340173721313,
2640
+ 0.6777340173721313
2641
+ ],
2642
+ [
2643
+ 1.25,
2644
+ 0.8085939884185791,
2645
+ 0.8085939884185791
2646
+ ],
2647
+ [
2648
+ 1.25,
2649
+ 0.8457030057907104,
2650
+ 0.8457030057907104
2651
+ ],
2652
+ [
2653
+ 1.25,
2654
+ 0.5859379768371582,
2655
+ 0.5859379768371582
2656
+ ],
2657
+ [
2658
+ 1.0,
2659
+ 0.732421875,
2660
+ 0.732421875
2661
+ ],
2662
+ [
2663
+ 1.25,
2664
+ 0.703125,
2665
+ 0.703125
2666
+ ],
2667
+ [
2668
+ 1.0,
2669
+ 0.732421875,
2670
+ 0.732421875
2671
+ ],
2672
+ [
2673
+ 1.0,
2674
+ 0.6484375,
2675
+ 0.6484375
2676
+ ],
2677
+ [
2678
+ 1.0,
2679
+ 0.748046875,
2680
+ 0.748046875
2681
+ ],
2682
+ [
2683
+ 1.25,
2684
+ 0.8085939884185791,
2685
+ 0.8085939884185791
2686
+ ],
2687
+ [
2688
+ 1.0,
2689
+ 0.74609375,
2690
+ 0.74609375
2691
+ ],
2692
+ [
2693
+ 1.25,
2694
+ 0.765625,
2695
+ 0.765625
2696
+ ],
2697
+ [
2698
+ 1.25,
2699
+ 0.7539060115814209,
2700
+ 0.7539060115814209
2701
+ ],
2702
+ [
2703
+ 1.25,
2704
+ 0.703125,
2705
+ 0.703125
2706
+ ],
2707
+ [
2708
+ 1.0,
2709
+ 0.8046875,
2710
+ 0.8046875
2711
+ ],
2712
+ [
2713
+ 1.25,
2714
+ 0.703125,
2715
+ 0.703125
2716
+ ],
2717
+ [
2718
+ 1.25,
2719
+ 0.7753909826278687,
2720
+ 0.7753909826278687
2721
+ ],
2722
+ [
2723
+ 1.25,
2724
+ 0.9042969942092896,
2725
+ 0.9042969942092896
2726
+ ],
2727
+ [
2728
+ 1.25,
2729
+ 0.9765620231628418,
2730
+ 0.9765620231628418
2731
+ ],
2732
+ [
2733
+ 1.0,
2734
+ 0.73828125,
2735
+ 0.73828125
2736
+ ],
2737
+ [
2738
+ 1.0,
2739
+ 0.830078125,
2740
+ 0.830078125
2741
+ ],
2742
+ [
2743
+ 1.0,
2744
+ 0.603515625,
2745
+ 0.603515625
2746
+ ],
2747
+ [
2748
+ 1.25,
2749
+ 0.8417969942092896,
2750
+ 0.8417969942092896
2751
+ ],
2752
+ [
2753
+ 1.25,
2754
+ 0.7714840173721313,
2755
+ 0.7714840173721313
2756
+ ],
2757
+ [
2758
+ 1.25,
2759
+ 0.7421879768371582,
2760
+ 0.7421879768371582
2761
+ ],
2762
+ [
2763
+ 1.25,
2764
+ 0.703125,
2765
+ 0.703125
2766
+ ],
2767
+ [
2768
+ 1.25,
2769
+ 0.84375,
2770
+ 0.84375
2771
+ ],
2772
+ [
2773
+ 1.25,
2774
+ 0.9550780057907104,
2775
+ 0.9550780057907104
2776
+ ],
2777
+ [
2778
+ 1.0,
2779
+ 0.681640625,
2780
+ 0.681640625
2781
+ ],
2782
+ [
2783
+ 1.0,
2784
+ 0.83984375,
2785
+ 0.83984375
2786
+ ],
2787
+ [
2788
+ 1.25,
2789
+ 0.8691409826278687,
2790
+ 0.8691409826278687
2791
+ ],
2792
+ [
2793
+ 1.0,
2794
+ 0.955078125,
2795
+ 0.955078125
2796
+ ],
2797
+ [
2798
+ 1.25,
2799
+ 0.8613280057907104,
2800
+ 0.8613280057907104
2801
+ ],
2802
+ [
2803
+ 1.0,
2804
+ 0.7109375,
2805
+ 0.7109375
2806
+ ],
2807
+ [
2808
+ 1.25,
2809
+ 0.765625,
2810
+ 0.765625
2811
+ ],
2812
+ [
2813
+ 1.0,
2814
+ 0.6875,
2815
+ 0.6875
2816
+ ],
2817
+ [
2818
+ 1.0,
2819
+ 0.775390625,
2820
+ 0.775390625
2821
+ ],
2822
+ [
2823
+ 1.25,
2824
+ 0.7558590173721313,
2825
+ 0.7558590173721313
2826
+ ],
2827
+ [
2828
+ 1.25,
2829
+ 0.6503909826278687,
2830
+ 0.6503909826278687
2831
+ ],
2832
+ [
2833
+ 1.0,
2834
+ 0.7109375,
2835
+ 0.7109375
2836
+ ],
2837
+ [
2838
+ 1.25,
2839
+ 0.859375,
2840
+ 0.859375
2841
+ ],
2842
+ [
2843
+ 1.0,
2844
+ 0.712890625,
2845
+ 0.712890625
2846
+ ],
2847
+ [
2848
+ 1.25,
2849
+ 0.8964840173721313,
2850
+ 0.8964840173721313
2851
+ ],
2852
+ [
2853
+ 1.25,
2854
+ 0.765625,
2855
+ 0.765625
2856
+ ],
2857
+ [
2858
+ 1.25,
2859
+ 0.703125,
2860
+ 0.703125
2861
+ ],
2862
+ [
2863
+ 1.25,
2864
+ 0.703125,
2865
+ 0.703125
2866
+ ],
2867
+ [
2868
+ 1.0,
2869
+ 0.62890625,
2870
+ 0.62890625
2871
+ ],
2872
+ [
2873
+ 1.25,
2874
+ 0.703125,
2875
+ 0.703125
2876
+ ],
2877
+ [
2878
+ 1.25,
2879
+ 0.9765620231628418,
2880
+ 0.9765620231628418
2881
+ ],
2882
+ [
2883
+ 1.0,
2884
+ 0.75390625,
2885
+ 0.75390625
2886
+ ],
2887
+ [
2888
+ 1.25,
2889
+ 0.7265620231628418,
2890
+ 0.7265620231628418
2891
+ ],
2892
+ [
2893
+ 1.25,
2894
+ 0.703125,
2895
+ 0.703125
2896
+ ],
2897
+ [
2898
+ 1.25,
2899
+ 0.7792969942092896,
2900
+ 0.7792969942092896
2901
+ ],
2902
+ [
2903
+ 1.0,
2904
+ 0.662109375,
2905
+ 0.662109375
2906
+ ],
2907
+ [
2908
+ 1.25,
2909
+ 0.703125,
2910
+ 0.703125
2911
+ ],
2912
+ [
2913
+ 1.25,
2914
+ 0.8359379768371582,
2915
+ 0.8359379768371582
2916
+ ],
2917
+ [
2918
+ 1.0,
2919
+ 0.8125,
2920
+ 0.8125
2921
+ ],
2922
+ [
2923
+ 1.0,
2924
+ 0.732421875,
2925
+ 0.732421875
2926
+ ],
2927
+ [
2928
+ 1.25,
2929
+ 0.5957030057907104,
2930
+ 0.5957030057907104
2931
+ ],
2932
+ [
2933
+ 1.0,
2934
+ 0.771484375,
2935
+ 0.771484375
2936
+ ],
2937
+ [
2938
+ 1.0,
2939
+ 0.759765625,
2940
+ 0.759765625
2941
+ ],
2942
+ [
2943
+ 1.25,
2944
+ 0.7636719942092896,
2945
+ 0.7636719942092896
2946
+ ],
2947
+ [
2948
+ 1.0,
2949
+ 0.921875,
2950
+ 0.921875
2951
+ ],
2952
+ [
2953
+ 1.25,
2954
+ 0.8027340173721313,
2955
+ 0.8027340173721313
2956
+ ],
2957
+ [
2958
+ 1.25,
2959
+ 0.7441409826278687,
2960
+ 0.7441409826278687
2961
+ ],
2962
+ [
2963
+ 1.25,
2964
+ 0.6796879768371582,
2965
+ 0.6796879768371582
2966
+ ],
2967
+ [
2968
+ 1.0,
2969
+ 0.732421875,
2970
+ 0.732421875
2971
+ ],
2972
+ [
2973
+ 1.25,
2974
+ 0.8359379768371582,
2975
+ 0.8359379768371582
2976
+ ],
2977
+ [
2978
+ 1.25,
2979
+ 0.703125,
2980
+ 0.703125
2981
+ ],
2982
+ [
2983
+ 1.0,
2984
+ 0.73828125,
2985
+ 0.73828125
2986
+ ],
2987
+ [
2988
+ 1.25,
2989
+ 0.7421879768371582,
2990
+ 0.7421879768371582
2991
+ ],
2992
+ [
2993
+ 1.25,
2994
+ 0.7578120231628418,
2995
+ 0.7578120231628418
2996
+ ],
2997
+ [
2998
+ 1.0,
2999
+ 0.58203125,
3000
+ 0.58203125
3001
+ ],
3002
+ [
3003
+ 1.0,
3004
+ 0.66015625,
3005
+ 0.66015625
3006
+ ],
3007
+ [
3008
+ 1.25,
3009
+ 0.6171879768371582,
3010
+ 0.6171879768371582
3011
+ ],
3012
+ [
3013
+ 1.25,
3014
+ 0.8105469942092896,
3015
+ 0.8105469942092896
3016
+ ],
3017
+ [
3018
+ 1.25,
3019
+ 0.7890620231628418,
3020
+ 0.7890620231628418
3021
+ ],
3022
+ [
3023
+ 1.25,
3024
+ 0.9765620231628418,
3025
+ 0.9765620231628418
3026
+ ],
3027
+ [
3028
+ 1.0,
3029
+ 0.693359375,
3030
+ 0.693359375
3031
+ ],
3032
+ [
3033
+ 1.25,
3034
+ 0.7636719942092896,
3035
+ 0.7636719942092896
3036
+ ],
3037
+ [
3038
+ 1.0,
3039
+ 0.76171875,
3040
+ 0.76171875
3041
+ ],
3042
+ [
3043
+ 1.25,
3044
+ 0.9453120231628418,
3045
+ 0.9453120231628418
3046
+ ],
3047
+ [
3048
+ 1.0,
3049
+ 0.642578125,
3050
+ 0.642578125
3051
+ ],
3052
+ [
3053
+ 1.0,
3054
+ 0.576171875,
3055
+ 0.576171875
3056
+ ],
3057
+ [
3058
+ 1.0,
3059
+ 0.6875,
3060
+ 0.6875
3061
+ ],
3062
+ [
3063
+ 1.25,
3064
+ 0.7304689884185791,
3065
+ 0.7304689884185791
3066
+ ],
3067
+ [
3068
+ 1.0,
3069
+ 0.71875,
3070
+ 0.71875
3071
+ ],
3072
+ [
3073
+ 1.25,
3074
+ 0.9414060115814209,
3075
+ 0.9414060115814209
3076
+ ],
3077
+ [
3078
+ 1.0,
3079
+ 0.720703125,
3080
+ 0.720703125
3081
+ ],
3082
+ [
3083
+ 1.25,
3084
+ 0.7519530057907104,
3085
+ 0.7519530057907104
3086
+ ],
3087
+ [
3088
+ 1.0,
3089
+ 0.587890625,
3090
+ 0.587890625
3091
+ ],
3092
+ [
3093
+ 1.25,
3094
+ 0.6503909826278687,
3095
+ 0.6503909826278687
3096
+ ],
3097
+ [
3098
+ 0.625,
3099
+ 0.75,
3100
+ 0.75
3101
+ ],
3102
+ [
3103
+ 1.25,
3104
+ 0.7832030057907104,
3105
+ 0.7832030057907104
3106
+ ],
3107
+ [
3108
+ 1.0,
3109
+ 0.7578125,
3110
+ 0.7578125
3111
+ ],
3112
+ [
3113
+ 1.25,
3114
+ 0.703125,
3115
+ 0.703125
3116
+ ],
3117
+ [
3118
+ 1.25,
3119
+ 0.9511719942092896,
3120
+ 0.9511719942092896
3121
+ ],
3122
+ [
3123
+ 1.0,
3124
+ 0.66796875,
3125
+ 0.66796875
3126
+ ],
3127
+ [
3128
+ 1.25,
3129
+ 0.7714840173721313,
3130
+ 0.7714840173721313
3131
+ ],
3132
+ [
3133
+ 1.0,
3134
+ 0.763671875,
3135
+ 0.763671875
3136
+ ],
3137
+ [
3138
+ 1.25,
3139
+ 0.9628909826278687,
3140
+ 0.9628909826278687
3141
+ ],
3142
+ [
3143
+ 1.0,
3144
+ 0.74609375,
3145
+ 0.74609375
3146
+ ],
3147
+ [
3148
+ 1.25,
3149
+ 0.8847659826278687,
3150
+ 0.8847659826278687
3151
+ ],
3152
+ [
3153
+ 1.25,
3154
+ 0.8066409826278687,
3155
+ 0.8066409826278687
3156
+ ],
3157
+ [
3158
+ 1.0,
3159
+ 0.685546875,
3160
+ 0.685546875
3161
+ ],
3162
+ [
3163
+ 1.25,
3164
+ 0.703125,
3165
+ 0.703125
3166
+ ],
3167
+ [
3168
+ 1.25,
3169
+ 0.8085939884185791,
3170
+ 0.8085939884185791
3171
+ ],
3172
+ [
3173
+ 1.0,
3174
+ 0.74609375,
3175
+ 0.74609375
3176
+ ],
3177
+ [
3178
+ 1.25,
3179
+ 0.6738280057907104,
3180
+ 0.6738280057907104
3181
+ ],
3182
+ [
3183
+ 1.25,
3184
+ 0.9023439884185791,
3185
+ 0.9023439884185791
3186
+ ],
3187
+ [
3188
+ 1.0,
3189
+ 0.767578125,
3190
+ 0.767578125
3191
+ ],
3192
+ [
3193
+ 1.0,
3194
+ 0.74609375,
3195
+ 0.74609375
3196
+ ],
3197
+ [
3198
+ 1.0,
3199
+ 0.615234375,
3200
+ 0.615234375
3201
+ ],
3202
+ [
3203
+ 1.25,
3204
+ 0.7421879768371582,
3205
+ 0.7421879768371582
3206
+ ],
3207
+ [
3208
+ 1.25,
3209
+ 0.7265620231628418,
3210
+ 0.7265620231628418
3211
+ ],
3212
+ [
3213
+ 1.25,
3214
+ 0.7773439884185791,
3215
+ 0.7773439884185791
3216
+ ],
3217
+ [
3218
+ 1.25,
3219
+ 0.8320310115814209,
3220
+ 0.8320310115814209
3221
+ ],
3222
+ [
3223
+ 1.25,
3224
+ 0.703125,
3225
+ 0.703125
3226
+ ],
3227
+ [
3228
+ 1.25,
3229
+ 0.765625,
3230
+ 0.765625
3231
+ ],
3232
+ [
3233
+ 1.25,
3234
+ 0.703125,
3235
+ 0.703125
3236
+ ],
3237
+ [
3238
+ 1.25,
3239
+ 0.7773439884185791,
3240
+ 0.7773439884185791
3241
+ ],
3242
+ [
3243
+ 1.25,
3244
+ 0.703125,
3245
+ 0.703125
3246
+ ],
3247
+ [
3248
+ 1.25,
3249
+ 0.6855469942092896,
3250
+ 0.6855469942092896
3251
+ ],
3252
+ [
3253
+ 1.0,
3254
+ 0.693359375,
3255
+ 0.693359375
3256
+ ],
3257
+ [
3258
+ 1.25,
3259
+ 0.8242189884185791,
3260
+ 0.8242189884185791
3261
+ ],
3262
+ [
3263
+ 1.0,
3264
+ 0.59765625,
3265
+ 0.59765625
3266
+ ],
3267
+ [
3268
+ 1.0,
3269
+ 0.947265625,
3270
+ 0.947265625
3271
+ ],
3272
+ [
3273
+ 1.0,
3274
+ 0.640625,
3275
+ 0.640625
3276
+ ],
3277
+ [
3278
+ 1.0,
3279
+ 0.7265625,
3280
+ 0.7265625
3281
+ ],
3282
+ [
3283
+ 1.25,
3284
+ 0.8046879768371582,
3285
+ 0.8046879768371582
3286
+ ],
3287
+ [
3288
+ 1.0,
3289
+ 0.833984375,
3290
+ 0.833984375
3291
+ ],
3292
+ [
3293
+ 1.0,
3294
+ 0.7265625,
3295
+ 0.7265625
3296
+ ],
3297
+ [
3298
+ 1.25,
3299
+ 0.7480469942092896,
3300
+ 0.7480469942092896
3301
+ ],
3302
+ [
3303
+ 1.25,
3304
+ 0.7050780057907104,
3305
+ 0.7050780057907104
3306
+ ],
3307
+ [
3308
+ 1.0,
3309
+ 0.66796875,
3310
+ 0.66796875
3311
+ ],
3312
+ [
3313
+ 1.25,
3314
+ 0.7753909826278687,
3315
+ 0.7753909826278687
3316
+ ],
3317
+ [
3318
+ 1.25,
3319
+ 0.7832030057907104,
3320
+ 0.7832030057907104
3321
+ ],
3322
+ [
3323
+ 1.25,
3324
+ 0.7304689884185791,
3325
+ 0.7304689884185791
3326
+ ],
3327
+ [
3328
+ 1.25,
3329
+ 0.7734379768371582,
3330
+ 0.7734379768371582
3331
+ ],
3332
+ [
3333
+ 1.25,
3334
+ 0.6894530057907104,
3335
+ 0.6894530057907104
3336
+ ],
3337
+ [
3338
+ 1.0,
3339
+ 0.814453125,
3340
+ 0.814453125
3341
+ ],
3342
+ [
3343
+ 1.25,
3344
+ 0.8964840173721313,
3345
+ 0.8964840173721313
3346
+ ],
3347
+ [
3348
+ 1.0,
3349
+ 0.646484375,
3350
+ 0.646484375
3351
+ ],
3352
+ [
3353
+ 1.0,
3354
+ 0.66015625,
3355
+ 0.66015625
3356
+ ],
3357
+ [
3358
+ 1.25,
3359
+ 0.8203120231628418,
3360
+ 0.8203120231628418
3361
+ ],
3362
+ [
3363
+ 1.25,
3364
+ 0.7890620231628418,
3365
+ 0.7890620231628418
3366
+ ],
3367
+ [
3368
+ 1.0,
3369
+ 0.87890625,
3370
+ 0.87890625
3371
+ ],
3372
+ [
3373
+ 1.0,
3374
+ 0.8515625,
3375
+ 0.8515625
3376
+ ],
3377
+ [
3378
+ 1.25,
3379
+ 0.9765620231628418,
3380
+ 0.9765620231628418
3381
+ ],
3382
+ [
3383
+ 1.25,
3384
+ 0.65625,
3385
+ 0.65625
3386
+ ],
3387
+ [
3388
+ 1.0,
3389
+ 0.6875,
3390
+ 0.6875
3391
+ ],
3392
+ [
3393
+ 1.25,
3394
+ 0.7382810115814209,
3395
+ 0.7382810115814209
3396
+ ],
3397
+ [
3398
+ 1.2499973773956299,
3399
+ 0.8632810115814209,
3400
+ 0.8632810115814209
3401
+ ],
3402
+ [
3403
+ 1.0,
3404
+ 0.732421875,
3405
+ 0.732421875
3406
+ ],
3407
+ [
3408
+ 1.25,
3409
+ 0.75,
3410
+ 0.75
3411
+ ],
3412
+ [
3413
+ 1.0,
3414
+ 0.90625,
3415
+ 0.90625
3416
+ ],
3417
+ [
3418
+ 1.0,
3419
+ 0.66796875,
3420
+ 0.66796875
3421
+ ],
3422
+ [
3423
+ 1.0,
3424
+ 0.783203125,
3425
+ 0.783203125
3426
+ ],
3427
+ [
3428
+ 1.25,
3429
+ 0.7441409826278687,
3430
+ 0.7441409826278687
3431
+ ],
3432
+ [
3433
+ 1.25,
3434
+ 0.7246090173721313,
3435
+ 0.7246090173721313
3436
+ ],
3437
+ [
3438
+ 1.25,
3439
+ 0.703125,
3440
+ 0.703125
3441
+ ],
3442
+ [
3443
+ 1.25,
3444
+ 0.7597659826278687,
3445
+ 0.7597659826278687
3446
+ ],
3447
+ [
3448
+ 1.0,
3449
+ 0.8359375,
3450
+ 0.8359375
3451
+ ],
3452
+ [
3453
+ 1.25,
3454
+ 0.7597659826278687,
3455
+ 0.7597659826278687
3456
+ ],
3457
+ [
3458
+ 1.25,
3459
+ 0.7988280057907104,
3460
+ 0.7988280057907104
3461
+ ],
3462
+ [
3463
+ 1.0,
3464
+ 0.794921875,
3465
+ 0.794921875
3466
+ ],
3467
+ [
3468
+ 1.0,
3469
+ 0.673828125,
3470
+ 0.673828125
3471
+ ],
3472
+ [
3473
+ 1.0,
3474
+ 0.619140625,
3475
+ 0.619140625
3476
+ ],
3477
+ [
3478
+ 1.25,
3479
+ 0.6738280057907104,
3480
+ 0.6738280057907104
3481
+ ],
3482
+ [
3483
+ 1.25,
3484
+ 0.7871090173721313,
3485
+ 0.7871090173721313
3486
+ ],
3487
+ [
3488
+ 1.0,
3489
+ 0.62890625,
3490
+ 0.62890625
3491
+ ],
3492
+ [
3493
+ 1.25,
3494
+ 0.8476560115814209,
3495
+ 0.8476560115814209
3496
+ ],
3497
+ [
3498
+ 1.25,
3499
+ 0.8671879768371582,
3500
+ 0.8671879768371582
3501
+ ],
3502
+ [
3503
+ 1.0,
3504
+ 0.83984375,
3505
+ 0.83984375
3506
+ ],
3507
+ [
3508
+ 1.25,
3509
+ 0.7578120231628418,
3510
+ 0.7578120231628418
3511
+ ],
3512
+ [
3513
+ 1.25,
3514
+ 0.7597659826278687,
3515
+ 0.7597659826278687
3516
+ ],
3517
+ [
3518
+ 1.25,
3519
+ 0.7226560115814209,
3520
+ 0.7226560115814209
3521
+ ],
3522
+ [
3523
+ 1.25,
3524
+ 0.7480469942092896,
3525
+ 0.7480469942092896
3526
+ ],
3527
+ [
3528
+ 1.0,
3529
+ 0.75,
3530
+ 0.75
3531
+ ],
3532
+ [
3533
+ 0.6200000047683716,
3534
+ 0.78125,
3535
+ 0.78125
3536
+ ],
3537
+ [
3538
+ 1.25,
3539
+ 0.8789060115814209,
3540
+ 0.8789060115814209
3541
+ ],
3542
+ [
3543
+ 1.0,
3544
+ 0.84375,
3545
+ 0.84375
3546
+ ],
3547
+ [
3548
+ 1.0,
3549
+ 0.70703125,
3550
+ 0.70703125
3551
+ ],
3552
+ [
3553
+ 1.25,
3554
+ 0.703125,
3555
+ 0.703125
3556
+ ],
3557
+ [
3558
+ 1.25,
3559
+ 0.7441409826278687,
3560
+ 0.7441409826278687
3561
+ ],
3562
+ [
3563
+ 1.0,
3564
+ 0.744140625,
3565
+ 0.744140625
3566
+ ],
3567
+ [
3568
+ 1.25,
3569
+ 0.7695310115814209,
3570
+ 0.7695310115814209
3571
+ ],
3572
+ [
3573
+ 1.0,
3574
+ 0.693359375,
3575
+ 0.693359375
3576
+ ],
3577
+ [
3578
+ 1.25,
3579
+ 0.7109379768371582,
3580
+ 0.7109379768371582
3581
+ ],
3582
+ [
3583
+ 1.0,
3584
+ 0.96484375,
3585
+ 0.96484375
3586
+ ],
3587
+ [
3588
+ 1.25,
3589
+ 0.9433590173721313,
3590
+ 0.9433590173721313
3591
+ ],
3592
+ [
3593
+ 1.25,
3594
+ 0.6347659826278687,
3595
+ 0.6347659826278687
3596
+ ],
3597
+ [
3598
+ 1.25,
3599
+ 0.8789060115814209,
3600
+ 0.8789060115814209
3601
+ ],
3602
+ [
3603
+ 1.25,
3604
+ 0.7148439884185791,
3605
+ 0.7148439884185791
3606
+ ],
3607
+ [
3608
+ 1.0,
3609
+ 0.6875,
3610
+ 0.6875
3611
+ ],
3612
+ [
3613
+ 3.75,
3614
+ 0.5761719942092896,
3615
+ 0.5761719942092896
3616
+ ],
3617
+ [
3618
+ 1.25,
3619
+ 0.703125,
3620
+ 0.703125
3621
+ ],
3622
+ [
3623
+ 1.0,
3624
+ 0.759765625,
3625
+ 0.759765625
3626
+ ],
3627
+ [
3628
+ 1.25,
3629
+ 0.8730469942092896,
3630
+ 0.8730469942092896
3631
+ ],
3632
+ [
3633
+ 1.0,
3634
+ 0.73828125,
3635
+ 0.73828125
3636
+ ],
3637
+ [
3638
+ 1.25,
3639
+ 0.703125,
3640
+ 0.703125
3641
+ ],
3642
+ [
3643
+ 1.25,
3644
+ 0.8359379768371582,
3645
+ 0.8359379768371582
3646
+ ],
3647
+ [
3648
+ 1.25,
3649
+ 0.75,
3650
+ 0.75
3651
+ ],
3652
+ [
3653
+ 1.25,
3654
+ 0.7636719942092896,
3655
+ 0.7636719942092896
3656
+ ],
3657
+ [
3658
+ 1.25,
3659
+ 0.703125,
3660
+ 0.703125
3661
+ ],
3662
+ [
3663
+ 1.0,
3664
+ 0.90234375,
3665
+ 0.90234375
3666
+ ],
3667
+ [
3668
+ 1.25,
3669
+ 0.6542969942092896,
3670
+ 0.6542969942092896
3671
+ ],
3672
+ [
3673
+ 1.0,
3674
+ 0.705078125,
3675
+ 0.705078125
3676
+ ],
3677
+ [
3678
+ 1.0,
3679
+ 0.830078125,
3680
+ 0.830078125
3681
+ ],
3682
+ [
3683
+ 1.0,
3684
+ 0.744140625,
3685
+ 0.744140625
3686
+ ],
3687
+ [
3688
+ 1.25,
3689
+ 0.8925780057907104,
3690
+ 0.8925780057907104
3691
+ ],
3692
+ [
3693
+ 5.0,
3694
+ 0.65625,
3695
+ 0.65625
3696
+ ],
3697
+ [
3698
+ 1.25,
3699
+ 0.640625,
3700
+ 0.640625
3701
+ ],
3702
+ [
3703
+ 1.25,
3704
+ 0.7441409826278687,
3705
+ 0.7441409826278687
3706
+ ],
3707
+ [
3708
+ 1.25,
3709
+ 0.6445310115814209,
3710
+ 0.6445310115814209
3711
+ ],
3712
+ [
3713
+ 1.25,
3714
+ 0.7636719942092896,
3715
+ 0.7636719942092896
3716
+ ],
3717
+ [
3718
+ 1.25,
3719
+ 0.78125,
3720
+ 0.78125
3721
+ ],
3722
+ [
3723
+ 1.25,
3724
+ 0.703125,
3725
+ 0.703125
3726
+ ],
3727
+ [
3728
+ 1.0,
3729
+ 0.783203125,
3730
+ 0.783203125
3731
+ ],
3732
+ [
3733
+ 1.0,
3734
+ 0.884765625,
3735
+ 0.884765625
3736
+ ],
3737
+ [
3738
+ 1.25,
3739
+ 0.703125,
3740
+ 0.703125
3741
+ ],
3742
+ [
3743
+ 1.0,
3744
+ 0.81640625,
3745
+ 0.81640625
3746
+ ],
3747
+ [
3748
+ 1.0,
3749
+ 0.763671875,
3750
+ 0.763671875
3751
+ ],
3752
+ [
3753
+ 1.0,
3754
+ 0.94921875,
3755
+ 0.94921875
3756
+ ],
3757
+ [
3758
+ 1.25,
3759
+ 0.703125,
3760
+ 0.703125
3761
+ ],
3762
+ [
3763
+ 1.25,
3764
+ 0.6484379768371582,
3765
+ 0.6484379768371582
3766
+ ],
3767
+ [
3768
+ 1.0,
3769
+ 0.70703125,
3770
+ 0.70703125
3771
+ ],
3772
+ [
3773
+ 1.0,
3774
+ 0.6640625,
3775
+ 0.6640625
3776
+ ],
3777
+ [
3778
+ 1.25,
3779
+ 0.7109379768371582,
3780
+ 0.7109379768371582
3781
+ ],
3782
+ [
3783
+ 1.0,
3784
+ 0.8515625,
3785
+ 0.8515625
3786
+ ],
3787
+ [
3788
+ 1.25,
3789
+ 0.9765620231628418,
3790
+ 0.9765620231628418
3791
+ ],
3792
+ [
3793
+ 1.25,
3794
+ 0.7421879768371582,
3795
+ 0.7421879768371582
3796
+ ],
3797
+ [
3798
+ 1.25,
3799
+ 0.7558590173721313,
3800
+ 0.7558590173721313
3801
+ ],
3802
+ [
3803
+ 1.0,
3804
+ 0.798828125,
3805
+ 0.798828125
3806
+ ],
3807
+ [
3808
+ 1.25,
3809
+ 0.6542969942092896,
3810
+ 0.6542969942092896
3811
+ ],
3812
+ [
3813
+ 1.0,
3814
+ 0.69140625,
3815
+ 0.69140625
3816
+ ],
3817
+ [
3818
+ 1.25,
3819
+ 0.8007810115814209,
3820
+ 0.8007810115814209
3821
+ ],
3822
+ [
3823
+ 1.25,
3824
+ 0.7441409826278687,
3825
+ 0.7441409826278687
3826
+ ],
3827
+ [
3828
+ 1.25,
3829
+ 0.78125,
3830
+ 0.78125
3831
+ ],
3832
+ [
3833
+ 1.25,
3834
+ 0.8378909826278687,
3835
+ 0.8378909826278687
3836
+ ],
3837
+ [
3838
+ 1.25,
3839
+ 0.6855469942092896,
3840
+ 0.6855469942092896
3841
+ ],
3842
+ [
3843
+ 1.25,
3844
+ 0.6386719942092896,
3845
+ 0.6386719942092896
3846
+ ],
3847
+ [
3848
+ 1.0,
3849
+ 0.658203125,
3850
+ 0.658203125
3851
+ ],
3852
+ [
3853
+ 1.0,
3854
+ 0.728515625,
3855
+ 0.728515625
3856
+ ],
3857
+ [
3858
+ 1.0,
3859
+ 0.779296875,
3860
+ 0.779296875
3861
+ ],
3862
+ [
3863
+ 1.0,
3864
+ 0.7421875,
3865
+ 0.7421875
3866
+ ],
3867
+ [
3868
+ 1.25,
3869
+ 0.8027340173721313,
3870
+ 0.8027340173721313
3871
+ ],
3872
+ [
3873
+ 1.25,
3874
+ 0.8984379768371582,
3875
+ 0.8984379768371582
3876
+ ],
3877
+ [
3878
+ 1.25,
3879
+ 0.7324219942092896,
3880
+ 0.7324219942092896
3881
+ ],
3882
+ [
3883
+ 1.0,
3884
+ 0.583984375,
3885
+ 0.583984375
3886
+ ],
3887
+ [
3888
+ 1.25,
3889
+ 0.9765620231628418,
3890
+ 0.9765620231628418
3891
+ ],
3892
+ [
3893
+ 1.0,
3894
+ 0.908203125,
3895
+ 0.908203125
3896
+ ],
3897
+ [
3898
+ 1.0,
3899
+ 0.6328125,
3900
+ 0.6328125
3901
+ ],
3902
+ [
3903
+ 1.25,
3904
+ 0.7539060115814209,
3905
+ 0.7539060115814209
3906
+ ],
3907
+ [
3908
+ 1.25,
3909
+ 0.75,
3910
+ 0.75
3911
+ ],
3912
+ [
3913
+ 1.25,
3914
+ 0.703125,
3915
+ 0.703125
3916
+ ],
3917
+ [
3918
+ 1.25,
3919
+ 0.6542969942092896,
3920
+ 0.6542969942092896
3921
+ ],
3922
+ [
3923
+ 1.0,
3924
+ 0.626953125,
3925
+ 0.626953125
3926
+ ],
3927
+ [
3928
+ 1.25,
3929
+ 0.8476560115814209,
3930
+ 0.8476560115814209
3931
+ ],
3932
+ [
3933
+ 1.25,
3934
+ 0.7480469942092896,
3935
+ 0.7480469942092896
3936
+ ],
3937
+ [
3938
+ 1.0,
3939
+ 0.6796875,
3940
+ 0.6796875
3941
+ ],
3942
+ [
3943
+ 1.0,
3944
+ 0.615234375,
3945
+ 0.615234375
3946
+ ],
3947
+ [
3948
+ 1.25,
3949
+ 0.7402340173721313,
3950
+ 0.7402340173721313
3951
+ ],
3952
+ [
3953
+ 1.0,
3954
+ 0.755859375,
3955
+ 0.755859375
3956
+ ],
3957
+ [
3958
+ 1.0,
3959
+ 0.673828125,
3960
+ 0.673828125
3961
+ ],
3962
+ [
3963
+ 1.25,
3964
+ 0.703125,
3965
+ 0.703125
3966
+ ],
3967
+ [
3968
+ 1.0,
3969
+ 0.693359375,
3970
+ 0.693359375
3971
+ ],
3972
+ [
3973
+ 1.0,
3974
+ 0.646484375,
3975
+ 0.646484375
3976
+ ],
3977
+ [
3978
+ 1.0,
3979
+ 0.58984375,
3980
+ 0.58984375
3981
+ ],
3982
+ [
3983
+ 1.0,
3984
+ 0.86328125,
3985
+ 0.86328125
3986
+ ],
3987
+ [
3988
+ 1.0,
3989
+ 0.705078125,
3990
+ 0.705078125
3991
+ ],
3992
+ [
3993
+ 1.0,
3994
+ 0.708984375,
3995
+ 0.708984375
3996
+ ],
3997
+ [
3998
+ 1.25,
3999
+ 0.8105469942092896,
4000
+ 0.8105469942092896
4001
+ ],
4002
+ [
4003
+ 1.0,
4004
+ 0.734375,
4005
+ 0.734375
4006
+ ],
4007
+ [
4008
+ 1.0,
4009
+ 0.8984375,
4010
+ 0.8984375
4011
+ ],
4012
+ [
4013
+ 1.25,
4014
+ 0.7265620231628418,
4015
+ 0.7265620231628418
4016
+ ],
4017
+ [
4018
+ 1.0,
4019
+ 0.623046875,
4020
+ 0.623046875
4021
+ ],
4022
+ [
4023
+ 1.25,
4024
+ 0.8085939884185791,
4025
+ 0.8085939884185791
4026
+ ],
4027
+ [
4028
+ 1.25,
4029
+ 0.78125,
4030
+ 0.78125
4031
+ ],
4032
+ [
4033
+ 1.0,
4034
+ 0.837890625,
4035
+ 0.837890625
4036
+ ],
4037
+ [
4038
+ 1.25,
4039
+ 0.7792969942092896,
4040
+ 0.7792969942092896
4041
+ ],
4042
+ [
4043
+ 1.25,
4044
+ 0.8027340173721313,
4045
+ 0.8027340173721313
4046
+ ],
4047
+ [
4048
+ 1.0,
4049
+ 0.833984375,
4050
+ 0.833984375
4051
+ ],
4052
+ [
4053
+ 1.25,
4054
+ 0.703125,
4055
+ 0.703125
4056
+ ],
4057
+ [
4058
+ 1.0,
4059
+ 0.765625,
4060
+ 0.765625
4061
+ ],
4062
+ [
4063
+ 1.0,
4064
+ 0.638671875,
4065
+ 0.638671875
4066
+ ],
4067
+ [
4068
+ 1.25,
4069
+ 0.9492189884185791,
4070
+ 0.9492189884185791
4071
+ ],
4072
+ [
4073
+ 1.25,
4074
+ 0.5820310115814209,
4075
+ 0.5820310115814209
4076
+ ],
4077
+ [
4078
+ 1.25,
4079
+ 0.8867189884185791,
4080
+ 0.8867189884185791
4081
+ ],
4082
+ [
4083
+ 1.25,
4084
+ 0.7226560115814209,
4085
+ 0.7226560115814209
4086
+ ],
4087
+ [
4088
+ 1.25,
4089
+ 0.7910159826278687,
4090
+ 0.7910159826278687
4091
+ ],
4092
+ [
4093
+ 1.0,
4094
+ 0.626953125,
4095
+ 0.626953125
4096
+ ],
4097
+ [
4098
+ 1.0,
4099
+ 0.7578125,
4100
+ 0.7578125
4101
+ ],
4102
+ [
4103
+ 1.25,
4104
+ 0.7753909826278687,
4105
+ 0.7753909826278687
4106
+ ],
4107
+ [
4108
+ 1.25,
4109
+ 0.9023439884185791,
4110
+ 0.9023439884185791
4111
+ ],
4112
+ [
4113
+ 1.0,
4114
+ 0.740234375,
4115
+ 0.740234375
4116
+ ],
4117
+ [
4118
+ 1.25,
4119
+ 0.9160159826278687,
4120
+ 0.9160159826278687
4121
+ ],
4122
+ [
4123
+ 1.25,
4124
+ 0.7167969942092896,
4125
+ 0.7167969942092896
4126
+ ],
4127
+ [
4128
+ 1.0,
4129
+ 0.86328125,
4130
+ 0.86328125
4131
+ ],
4132
+ [
4133
+ 1.25,
4134
+ 0.7363280057907104,
4135
+ 0.7363280057907104
4136
+ ],
4137
+ [
4138
+ 1.0,
4139
+ 0.744140625,
4140
+ 0.744140625
4141
+ ],
4142
+ [
4143
+ 1.25,
4144
+ 0.7519530057907104,
4145
+ 0.7519530057907104
4146
+ ],
4147
+ [
4148
+ 1.25,
4149
+ 0.9765620231628418,
4150
+ 0.9765620231628418
4151
+ ],
4152
+ [
4153
+ 1.25,
4154
+ 0.7714840173721313,
4155
+ 0.7714840173721313
4156
+ ],
4157
+ [
4158
+ 1.25,
4159
+ 0.703125,
4160
+ 0.703125
4161
+ ],
4162
+ [
4163
+ 1.25,
4164
+ 0.8828120231628418,
4165
+ 0.8828120231628418
4166
+ ],
4167
+ [
4168
+ 1.0,
4169
+ 0.8359375,
4170
+ 0.8359375
4171
+ ],
4172
+ [
4173
+ 1.25,
4174
+ 0.7519530057907104,
4175
+ 0.7519530057907104
4176
+ ],
4177
+ [
4178
+ 1.25,
4179
+ 0.703125,
4180
+ 0.703125
4181
+ ],
4182
+ [
4183
+ 1.25,
4184
+ 0.8007810115814209,
4185
+ 0.8007810115814209
4186
+ ],
4187
+ [
4188
+ 1.25,
4189
+ 0.8789060115814209,
4190
+ 0.8789060115814209
4191
+ ],
4192
+ [
4193
+ 1.25,
4194
+ 0.703125,
4195
+ 0.703125
4196
+ ],
4197
+ [
4198
+ 1.0,
4199
+ 0.92578125,
4200
+ 0.92578125
4201
+ ],
4202
+ [
4203
+ 1.25,
4204
+ 0.6601560115814209,
4205
+ 0.6601560115814209
4206
+ ],
4207
+ [
4208
+ 1.0,
4209
+ 0.791015625,
4210
+ 0.791015625
4211
+ ],
4212
+ [
4213
+ 1.0,
4214
+ 0.96875,
4215
+ 0.96875
4216
+ ],
4217
+ [
4218
+ 1.25,
4219
+ 0.7871090173721313,
4220
+ 0.7871090173721313
4221
+ ],
4222
+ [
4223
+ 1.25,
4224
+ 0.8496090173721313,
4225
+ 0.8496090173721313
4226
+ ],
4227
+ [
4228
+ 1.0,
4229
+ 0.814453125,
4230
+ 0.814453125
4231
+ ],
4232
+ [
4233
+ 1.25,
4234
+ 0.7480469942092896,
4235
+ 0.7480469942092896
4236
+ ],
4237
+ [
4238
+ 1.25,
4239
+ 0.6582030057907104,
4240
+ 0.6582030057907104
4241
+ ],
4242
+ [
4243
+ 1.25,
4244
+ 0.703125,
4245
+ 0.703125
4246
+ ]
4247
+ ]
4248
+ }
Dataset307_radiologist_annotations_Merlin_Plus/nnUNetTrainer__nnUNetPlannerResEncL_torchres_isotropic__3d_fullres/fold_all/checkpoint_best.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:f83e25b4da9b2d5dc5cca5e766f480d08784f9633ef2933187159b4d08ed97b2
3
+ size 824287410
Dataset307_radiologist_annotations_Merlin_Plus/nnUNetTrainer__nnUNetPlannerResEncL_torchres_isotropic__3d_fullres/fold_all/checkpoint_final.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:4677ee02d0663a4111c147bba79049b3e01d6ac31f64e86491b8f9f06b9fda52
3
+ size 824292614
Dataset307_radiologist_annotations_Merlin_Plus/nnUNetTrainer__nnUNetPlannerResEncL_torchres_isotropic__3d_fullres/fold_all/debug.json ADDED
@@ -0,0 +1,53 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_best_ema": "None",
3
+ "batch_size": "2",
4
+ "configuration_manager": "{'data_identifier': 'nnUNetPlannerResEncL_torchres_isotropic_3d_fullres', 'preprocessor_name': 'DefaultPreprocessor', 'batch_size': 2, 'patch_size': [224, 160, 160], 'median_image_size_in_voxels': [469.0, 383.0, 383.0], 'spacing': [1.0, 1.0, 1.0], 'normalization_schemes': ['CTNormalization'], 'use_mask_for_norm': [False], 'resampling_fn_data': 'resample_torch_fornnunet', 'resampling_fn_seg': 'resample_torch_fornnunet', 'resampling_fn_data_kwargs': {'is_seg': False, 'force_separate_z': False, 'memefficient_seg_resampling': False}, 'resampling_fn_seg_kwargs': {'is_seg': True, 'force_separate_z': False, 'memefficient_seg_resampling': False}, 'resampling_fn_probabilities': 'resample_torch_fornnunet', 'resampling_fn_probabilities_kwargs': {'is_seg': False, 'force_separate_z': False, 'memefficient_seg_resampling': False}, 'architecture': {'network_class_name': 'dynamic_network_architectures.architectures.unet.ResidualEncoderUNet', 'arch_kwargs': {'n_stages': 6, 'features_per_stage': [32, 64, 128, 256, 320, 320], 'conv_op': 'torch.nn.modules.conv.Conv3d', 'kernel_sizes': [[3, 3, 3], [3, 3, 3], [3, 3, 3], [3, 3, 3], [3, 3, 3], [3, 3, 3]], 'strides': [[1, 1, 1], [2, 2, 2], [2, 2, 2], [2, 2, 2], [2, 2, 2], [2, 2, 2]], 'n_blocks_per_stage': [1, 3, 4, 6, 6, 6], 'n_conv_per_stage_decoder': [1, 1, 1, 1, 1], 'conv_bias': True, 'norm_op': 'torch.nn.modules.instancenorm.InstanceNorm3d', 'norm_op_kwargs': {'eps': 1e-05, 'affine': True}, 'dropout_op': None, 'dropout_op_kwargs': None, 'nonlin': 'torch.nn.LeakyReLU', 'nonlin_kwargs': {'inplace': True}}, '_kw_requires_import': ['conv_op', 'norm_op', 'dropout_op', 'nonlin']}, 'batch_dice': True}",
5
+ "configuration_name": "3d_fullres",
6
+ "cudnn_version": 90100,
7
+ "current_epoch": "0",
8
+ "dataloader_train": "<batchgenerators.dataloading.nondet_multi_threaded_augmenter.NonDetMultiThreadedAugmenter object at 0x155198adf610>",
9
+ "dataloader_train.generator": "<nnunetv2.training.dataloading.data_loader.nnUNetDataLoader object at 0x1553d76a00d0>",
10
+ "dataloader_train.num_processes": "12",
11
+ "dataloader_train.transform": "None",
12
+ "dataloader_val": "<batchgenerators.dataloading.nondet_multi_threaded_augmenter.NonDetMultiThreadedAugmenter object at 0x1553d76a02b0>",
13
+ "dataloader_val.generator": "<nnunetv2.training.dataloading.data_loader.nnUNetDataLoader object at 0x155198adf5e0>",
14
+ "dataloader_val.num_processes": "6",
15
+ "dataloader_val.transform": "None",
16
+ "dataset_json": "{'channel_names': {'0': 'CT'}, 'labels': {'background': 0, 'kidney_right': 1, 'kidney_left': 2, 'kidney_lesion': 3, 'kidney_lesion_kidney_right': 4, 'kidney_lesion_kidney_left': 5, 'pancreas': 6, 'pancreas_head': 7, 'pancreas_body': 8, 'pancreas_tail': 9, 'pancreatic_lesion': 10, 'pancreatic_lesion_pancreas_head': 11, 'pancreatic_lesion_pancreas_body': 12, 'pancreatic_lesion_pancreas_tail': 13, 'liver': 14, 'liver_segment_1': 15, 'liver_segment_2': 16, 'liver_segment_3': 17, 'liver_segment_4': 18, 'liver_segment_5': 19, 'liver_segment_6': 20, 'liver_segment_7': 21, 'liver_segment_8': 22, 'liver_lesion': 23, 'liver_lesion_liver_segment_1': 24, 'liver_lesion_liver_segment_2': 25, 'liver_lesion_liver_segment_3': 26, 'liver_lesion_liver_segment_4': 27, 'liver_lesion_liver_segment_5': 28, 'liver_lesion_liver_segment_6': 29, 'liver_lesion_liver_segment_7': 30, 'liver_lesion_liver_segment_8': 31, 'spleen': 32, 'spleen_lesion': 33, 'colon': 34, 'colon_lesion': 35, 'stomach': 36, 'stomach_lesion': 37, 'duodenum': 38, 'duodenum_lesion': 39, 'common_bile_duct': 40, 'intestine': 41, 'aorta': 42, 'postcava': 43, 'adrenal_gland_left': 44, 'adrenal_gland_right': 45, 'adrenal_lesion': 46, 'adrenal_lesion_adrenal_gland_left': 47, 'adrenal_lesion_adrenal_gland_right': 48, 'gall_bladder': 49, 'gallbladder_lesion': 50, 'gallbladder_lesion_gall_bladder': 51, 'bladder': 52, 'bladder_lesion': 53, 'celiac_trunk': 54, 'esophagus': 55, 'esophagus_lesion': 56, 'hepatic_vessel': 57, 'portal_vein_and_splenic_vein': 58, 'lung_left': 59, 'lung_right': 60, 'lung_lesion': 61, 'lung_lesion_lung_left': 62, 'lung_lesion_lung_right': 63, 'prostate': 64, 'prostate_lesion': 65, 'uterus_lesion': 66, 'uterus_lesion_prostate': 67, 'rectum': 68, 'femur_left': 69, 'femur_right': 70, 'superior_mesenteric_artery': 71, 'veins': 72, 'bone_lesion': 73, 'breast_lesion': 74}, 'numTraining': 423, 'file_ending': '.nii.gz', 'licence': \"Whoever converted this dataset was lazy and didn't look it up!\", 'converted_by': 'Please enter your name, especially when sharing datasets with others in a common infrastructure!', 'citation': None, 'overwrite_image_reader_writer': 'NibabelIOWithReorient', 'regions_class_order': 'Dataset307_radiologist_annotations_Merlin_Plus'}",
17
+ "device": "cuda:0",
18
+ "disable_checkpointing": "False",
19
+ "enable_deep_supervision": "True",
20
+ "fold": "all",
21
+ "folder_with_segs_from_previous_stage": "None",
22
+ "gpu_name": "NVIDIA H100 80GB HBM3",
23
+ "grad_scaler": "<torch.amp.grad_scaler.GradScaler object at 0x1553b7c2dea0>",
24
+ "hostname": "gpu312",
25
+ "inference_allowed_mirroring_axes": "(0, 1, 2)",
26
+ "initial_lr": "0.01",
27
+ "is_cascaded": "False",
28
+ "is_ddp": "False",
29
+ "label_manager": "<nnunetv2.utilities.label_handling.label_handling.LabelManager object at 0x1553b7c2dd50>",
30
+ "local_rank": "0",
31
+ "log_file": "/projects/bodymaps/Pedro/data/nnUNet_results/Dataset307_radiologist_annotations_Merlin_Plus/nnUNetTrainer__nnUNetPlannerResEncL_torchres_isotropic__3d_fullres/fold_all/training_log_2026_2_21_12_15_45.txt",
32
+ "logger": "<nnunetv2.training.logging.nnunet_logger.nnUNetLogger object at 0x1553b7c2dd20>",
33
+ "loss": "DeepSupervisionWrapper(\n (loss): DC_and_CE_loss(\n (ce): RobustCrossEntropyLoss()\n (dc): OptimizedModule(\n (_orig_mod): MemoryEfficientSoftDiceLoss()\n )\n )\n)",
34
+ "lr_scheduler": "<nnunetv2.training.lr_scheduler.polylr.PolyLRScheduler object at 0x1551aad33bb0>",
35
+ "my_init_kwargs": "{'plans': {'dataset_name': 'Dataset307_radiologist_annotations_Merlin_Plus', 'plans_name': 'nnUNetPlannerResEncL_torchres_isotropic', 'original_median_spacing_after_transp': [1.25, 0.748046875, 0.748046875], 'original_median_shape_after_transp': [409, 512, 512], 'image_reader_writer': 'NibabelIOWithReorient', 'transpose_forward': [0, 1, 2], 'transpose_backward': [0, 1, 2], 'configurations': {'2d': {'data_identifier': 'nnUNetPlannerResEncL_torchres_isotropic_2d', 'preprocessor_name': 'DefaultPreprocessor', 'batch_size': 55, 'patch_size': [384, 384], 'median_image_size_in_voxels': [383.0, 383.0], 'spacing': [1.0, 1.0], 'normalization_schemes': ['CTNormalization'], 'use_mask_for_norm': [False], 'resampling_fn_data': 'resample_torch_fornnunet', 'resampling_fn_seg': 'resample_torch_fornnunet', 'resampling_fn_data_kwargs': {'is_seg': False, 'force_separate_z': False, 'memefficient_seg_resampling': False}, 'resampling_fn_seg_kwargs': {'is_seg': True, 'force_separate_z': False, 'memefficient_seg_resampling': False}, 'resampling_fn_probabilities': 'resample_torch_fornnunet', 'resampling_fn_probabilities_kwargs': {'is_seg': False, 'force_separate_z': False, 'memefficient_seg_resampling': False}, 'architecture': {'network_class_name': 'dynamic_network_architectures.architectures.unet.ResidualEncoderUNet', 'arch_kwargs': {'n_stages': 7, 'features_per_stage': [32, 64, 128, 256, 512, 512, 512], 'conv_op': 'torch.nn.modules.conv.Conv2d', 'kernel_sizes': [[3, 3], [3, 3], [3, 3], [3, 3], [3, 3], [3, 3], [3, 3]], 'strides': [[1, 1], [2, 2], [2, 2], [2, 2], [2, 2], [2, 2], [2, 2]], 'n_blocks_per_stage': [1, 3, 4, 6, 6, 6, 6], 'n_conv_per_stage_decoder': [1, 1, 1, 1, 1, 1], 'conv_bias': True, 'norm_op': 'torch.nn.modules.instancenorm.InstanceNorm2d', 'norm_op_kwargs': {'eps': 1e-05, 'affine': True}, 'dropout_op': None, 'dropout_op_kwargs': None, 'nonlin': 'torch.nn.LeakyReLU', 'nonlin_kwargs': {'inplace': True}}, '_kw_requires_import': ['conv_op', 'norm_op', 'dropout_op', 'nonlin']}, 'batch_dice': True}, '3d_lowres': {'data_identifier': 'nnUNetPlannerResEncL_torchres_isotropic_3d_lowres', 'preprocessor_name': 'DefaultPreprocessor', 'batch_size': 2, 'patch_size': [224, 160, 160], 'median_image_size_in_voxels': [319, 261, 261], 'spacing': [1.4685337134515648, 1.4685337134515648, 1.4685337134515648], 'normalization_schemes': ['CTNormalization'], 'use_mask_for_norm': [False], 'resampling_fn_data': 'resample_torch_fornnunet', 'resampling_fn_seg': 'resample_torch_fornnunet', 'resampling_fn_data_kwargs': {'is_seg': False, 'force_separate_z': False, 'memefficient_seg_resampling': False}, 'resampling_fn_seg_kwargs': {'is_seg': True, 'force_separate_z': False, 'memefficient_seg_resampling': False}, 'resampling_fn_probabilities': 'resample_torch_fornnunet', 'resampling_fn_probabilities_kwargs': {'is_seg': False, 'force_separate_z': False, 'memefficient_seg_resampling': False}, 'architecture': {'network_class_name': 'dynamic_network_architectures.architectures.unet.ResidualEncoderUNet', 'arch_kwargs': {'n_stages': 6, 'features_per_stage': [32, 64, 128, 256, 320, 320], 'conv_op': 'torch.nn.modules.conv.Conv3d', 'kernel_sizes': [[3, 3, 3], [3, 3, 3], [3, 3, 3], [3, 3, 3], [3, 3, 3], [3, 3, 3]], 'strides': [[1, 1, 1], [2, 2, 2], [2, 2, 2], [2, 2, 2], [2, 2, 2], [2, 2, 2]], 'n_blocks_per_stage': [1, 3, 4, 6, 6, 6], 'n_conv_per_stage_decoder': [1, 1, 1, 1, 1], 'conv_bias': True, 'norm_op': 'torch.nn.modules.instancenorm.InstanceNorm3d', 'norm_op_kwargs': {'eps': 1e-05, 'affine': True}, 'dropout_op': None, 'dropout_op_kwargs': None, 'nonlin': 'torch.nn.LeakyReLU', 'nonlin_kwargs': {'inplace': True}}, '_kw_requires_import': ['conv_op', 'norm_op', 'dropout_op', 'nonlin']}, 'batch_dice': False, 'next_stage': '3d_cascade_fullres'}, '3d_fullres': {'data_identifier': 'nnUNetPlannerResEncL_torchres_isotropic_3d_fullres', 'preprocessor_name': 'DefaultPreprocessor', 'batch_size': 2, 'patch_size': [224, 160, 160], 'median_image_size_in_voxels': [469.0, 383.0, 383.0], 'spacing': [1.0, 1.0, 1.0], 'normalization_schemes': ['CTNormalization'], 'use_mask_for_norm': [False], 'resampling_fn_data': 'resample_torch_fornnunet', 'resampling_fn_seg': 'resample_torch_fornnunet', 'resampling_fn_data_kwargs': {'is_seg': False, 'force_separate_z': False, 'memefficient_seg_resampling': False}, 'resampling_fn_seg_kwargs': {'is_seg': True, 'force_separate_z': False, 'memefficient_seg_resampling': False}, 'resampling_fn_probabilities': 'resample_torch_fornnunet', 'resampling_fn_probabilities_kwargs': {'is_seg': False, 'force_separate_z': False, 'memefficient_seg_resampling': False}, 'architecture': {'network_class_name': 'dynamic_network_architectures.architectures.unet.ResidualEncoderUNet', 'arch_kwargs': {'n_stages': 6, 'features_per_stage': [32, 64, 128, 256, 320, 320], 'conv_op': 'torch.nn.modules.conv.Conv3d', 'kernel_sizes': [[3, 3, 3], [3, 3, 3], [3, 3, 3], [3, 3, 3], [3, 3, 3], [3, 3, 3]], 'strides': [[1, 1, 1], [2, 2, 2], [2, 2, 2], [2, 2, 2], [2, 2, 2], [2, 2, 2]], 'n_blocks_per_stage': [1, 3, 4, 6, 6, 6], 'n_conv_per_stage_decoder': [1, 1, 1, 1, 1], 'conv_bias': True, 'norm_op': 'torch.nn.modules.instancenorm.InstanceNorm3d', 'norm_op_kwargs': {'eps': 1e-05, 'affine': True}, 'dropout_op': None, 'dropout_op_kwargs': None, 'nonlin': 'torch.nn.LeakyReLU', 'nonlin_kwargs': {'inplace': True}}, '_kw_requires_import': ['conv_op', 'norm_op', 'dropout_op', 'nonlin']}, 'batch_dice': True}, '3d_cascade_fullres': {'inherits_from': '3d_fullres', 'previous_stage': '3d_lowres'}}, 'experiment_planner_used': 'nnUNetPlannerResEncL_torchres', 'label_manager': 'LabelManager', 'foreground_intensity_properties_per_channel': {'0': {'max': 1000.0, 'mean': -128.6943359375, 'median': 51.01091003417969, 'min': -1000.0, 'percentile_00_5': -996.0021362304688, 'percentile_99_5': 851.0109252929688, 'std': 398.4421691894531}}}, 'configuration': '3d_fullres', 'fold': 'all', 'dataset_json': {'channel_names': {'0': 'CT'}, 'labels': {'background': 0, 'kidney_right': 1, 'kidney_left': 2, 'kidney_lesion': 3, 'kidney_lesion_kidney_right': 4, 'kidney_lesion_kidney_left': 5, 'pancreas': 6, 'pancreas_head': 7, 'pancreas_body': 8, 'pancreas_tail': 9, 'pancreatic_lesion': 10, 'pancreatic_lesion_pancreas_head': 11, 'pancreatic_lesion_pancreas_body': 12, 'pancreatic_lesion_pancreas_tail': 13, 'liver': 14, 'liver_segment_1': 15, 'liver_segment_2': 16, 'liver_segment_3': 17, 'liver_segment_4': 18, 'liver_segment_5': 19, 'liver_segment_6': 20, 'liver_segment_7': 21, 'liver_segment_8': 22, 'liver_lesion': 23, 'liver_lesion_liver_segment_1': 24, 'liver_lesion_liver_segment_2': 25, 'liver_lesion_liver_segment_3': 26, 'liver_lesion_liver_segment_4': 27, 'liver_lesion_liver_segment_5': 28, 'liver_lesion_liver_segment_6': 29, 'liver_lesion_liver_segment_7': 30, 'liver_lesion_liver_segment_8': 31, 'spleen': 32, 'spleen_lesion': 33, 'colon': 34, 'colon_lesion': 35, 'stomach': 36, 'stomach_lesion': 37, 'duodenum': 38, 'duodenum_lesion': 39, 'common_bile_duct': 40, 'intestine': 41, 'aorta': 42, 'postcava': 43, 'adrenal_gland_left': 44, 'adrenal_gland_right': 45, 'adrenal_lesion': 46, 'adrenal_lesion_adrenal_gland_left': 47, 'adrenal_lesion_adrenal_gland_right': 48, 'gall_bladder': 49, 'gallbladder_lesion': 50, 'gallbladder_lesion_gall_bladder': 51, 'bladder': 52, 'bladder_lesion': 53, 'celiac_trunk': 54, 'esophagus': 55, 'esophagus_lesion': 56, 'hepatic_vessel': 57, 'portal_vein_and_splenic_vein': 58, 'lung_left': 59, 'lung_right': 60, 'lung_lesion': 61, 'lung_lesion_lung_left': 62, 'lung_lesion_lung_right': 63, 'prostate': 64, 'prostate_lesion': 65, 'uterus_lesion': 66, 'uterus_lesion_prostate': 67, 'rectum': 68, 'femur_left': 69, 'femur_right': 70, 'superior_mesenteric_artery': 71, 'veins': 72, 'bone_lesion': 73, 'breast_lesion': 74}, 'numTraining': 423, 'file_ending': '.nii.gz', 'licence': \"Whoever converted this dataset was lazy and didn't look it up!\", 'converted_by': 'Please enter your name, especially when sharing datasets with others in a common infrastructure!', 'citation': None, 'overwrite_image_reader_writer': 'NibabelIOWithReorient', 'regions_class_order': 'Dataset307_radiologist_annotations_Merlin_Plus'}, 'device': device(type='cuda')}",
36
+ "network": "OptimizedModule",
37
+ "num_epochs": "1000",
38
+ "num_input_channels": "1",
39
+ "num_iterations_per_epoch": "250",
40
+ "num_val_iterations_per_epoch": "50",
41
+ "optimizer": "SGD (\nParameter Group 0\n dampening: 0\n differentiable: False\n foreach: None\n fused: None\n initial_lr: 0.01\n lr: 0.01\n maximize: False\n momentum: 0.99\n nesterov: True\n weight_decay: 3e-05\n)",
42
+ "output_folder": "/projects/bodymaps/Pedro/data/nnUNet_results/Dataset307_radiologist_annotations_Merlin_Plus/nnUNetTrainer__nnUNetPlannerResEncL_torchres_isotropic__3d_fullres/fold_all",
43
+ "output_folder_base": "/projects/bodymaps/Pedro/data/nnUNet_results/Dataset307_radiologist_annotations_Merlin_Plus/nnUNetTrainer__nnUNetPlannerResEncL_torchres_isotropic__3d_fullres",
44
+ "oversample_foreground_percent": "0.33",
45
+ "plans_manager": "{'dataset_name': 'Dataset307_radiologist_annotations_Merlin_Plus', 'plans_name': 'nnUNetPlannerResEncL_torchres_isotropic', 'original_median_spacing_after_transp': [1.25, 0.748046875, 0.748046875], 'original_median_shape_after_transp': [409, 512, 512], 'image_reader_writer': 'NibabelIOWithReorient', 'transpose_forward': [0, 1, 2], 'transpose_backward': [0, 1, 2], 'configurations': {'2d': {'data_identifier': 'nnUNetPlannerResEncL_torchres_isotropic_2d', 'preprocessor_name': 'DefaultPreprocessor', 'batch_size': 55, 'patch_size': [384, 384], 'median_image_size_in_voxels': [383.0, 383.0], 'spacing': [1.0, 1.0], 'normalization_schemes': ['CTNormalization'], 'use_mask_for_norm': [False], 'resampling_fn_data': 'resample_torch_fornnunet', 'resampling_fn_seg': 'resample_torch_fornnunet', 'resampling_fn_data_kwargs': {'is_seg': False, 'force_separate_z': False, 'memefficient_seg_resampling': False}, 'resampling_fn_seg_kwargs': {'is_seg': True, 'force_separate_z': False, 'memefficient_seg_resampling': False}, 'resampling_fn_probabilities': 'resample_torch_fornnunet', 'resampling_fn_probabilities_kwargs': {'is_seg': False, 'force_separate_z': False, 'memefficient_seg_resampling': False}, 'architecture': {'network_class_name': 'dynamic_network_architectures.architectures.unet.ResidualEncoderUNet', 'arch_kwargs': {'n_stages': 7, 'features_per_stage': [32, 64, 128, 256, 512, 512, 512], 'conv_op': 'torch.nn.modules.conv.Conv2d', 'kernel_sizes': [[3, 3], [3, 3], [3, 3], [3, 3], [3, 3], [3, 3], [3, 3]], 'strides': [[1, 1], [2, 2], [2, 2], [2, 2], [2, 2], [2, 2], [2, 2]], 'n_blocks_per_stage': [1, 3, 4, 6, 6, 6, 6], 'n_conv_per_stage_decoder': [1, 1, 1, 1, 1, 1], 'conv_bias': True, 'norm_op': 'torch.nn.modules.instancenorm.InstanceNorm2d', 'norm_op_kwargs': {'eps': 1e-05, 'affine': True}, 'dropout_op': None, 'dropout_op_kwargs': None, 'nonlin': 'torch.nn.LeakyReLU', 'nonlin_kwargs': {'inplace': True}}, '_kw_requires_import': ['conv_op', 'norm_op', 'dropout_op', 'nonlin']}, 'batch_dice': True}, '3d_lowres': {'data_identifier': 'nnUNetPlannerResEncL_torchres_isotropic_3d_lowres', 'preprocessor_name': 'DefaultPreprocessor', 'batch_size': 2, 'patch_size': [224, 160, 160], 'median_image_size_in_voxels': [319, 261, 261], 'spacing': [1.4685337134515648, 1.4685337134515648, 1.4685337134515648], 'normalization_schemes': ['CTNormalization'], 'use_mask_for_norm': [False], 'resampling_fn_data': 'resample_torch_fornnunet', 'resampling_fn_seg': 'resample_torch_fornnunet', 'resampling_fn_data_kwargs': {'is_seg': False, 'force_separate_z': False, 'memefficient_seg_resampling': False}, 'resampling_fn_seg_kwargs': {'is_seg': True, 'force_separate_z': False, 'memefficient_seg_resampling': False}, 'resampling_fn_probabilities': 'resample_torch_fornnunet', 'resampling_fn_probabilities_kwargs': {'is_seg': False, 'force_separate_z': False, 'memefficient_seg_resampling': False}, 'architecture': {'network_class_name': 'dynamic_network_architectures.architectures.unet.ResidualEncoderUNet', 'arch_kwargs': {'n_stages': 6, 'features_per_stage': [32, 64, 128, 256, 320, 320], 'conv_op': 'torch.nn.modules.conv.Conv3d', 'kernel_sizes': [[3, 3, 3], [3, 3, 3], [3, 3, 3], [3, 3, 3], [3, 3, 3], [3, 3, 3]], 'strides': [[1, 1, 1], [2, 2, 2], [2, 2, 2], [2, 2, 2], [2, 2, 2], [2, 2, 2]], 'n_blocks_per_stage': [1, 3, 4, 6, 6, 6], 'n_conv_per_stage_decoder': [1, 1, 1, 1, 1], 'conv_bias': True, 'norm_op': 'torch.nn.modules.instancenorm.InstanceNorm3d', 'norm_op_kwargs': {'eps': 1e-05, 'affine': True}, 'dropout_op': None, 'dropout_op_kwargs': None, 'nonlin': 'torch.nn.LeakyReLU', 'nonlin_kwargs': {'inplace': True}}, '_kw_requires_import': ['conv_op', 'norm_op', 'dropout_op', 'nonlin']}, 'batch_dice': False, 'next_stage': '3d_cascade_fullres'}, '3d_fullres': {'data_identifier': 'nnUNetPlannerResEncL_torchres_isotropic_3d_fullres', 'preprocessor_name': 'DefaultPreprocessor', 'batch_size': 2, 'patch_size': [224, 160, 160], 'median_image_size_in_voxels': [469.0, 383.0, 383.0], 'spacing': [1.0, 1.0, 1.0], 'normalization_schemes': ['CTNormalization'], 'use_mask_for_norm': [False], 'resampling_fn_data': 'resample_torch_fornnunet', 'resampling_fn_seg': 'resample_torch_fornnunet', 'resampling_fn_data_kwargs': {'is_seg': False, 'force_separate_z': False, 'memefficient_seg_resampling': False}, 'resampling_fn_seg_kwargs': {'is_seg': True, 'force_separate_z': False, 'memefficient_seg_resampling': False}, 'resampling_fn_probabilities': 'resample_torch_fornnunet', 'resampling_fn_probabilities_kwargs': {'is_seg': False, 'force_separate_z': False, 'memefficient_seg_resampling': False}, 'architecture': {'network_class_name': 'dynamic_network_architectures.architectures.unet.ResidualEncoderUNet', 'arch_kwargs': {'n_stages': 6, 'features_per_stage': [32, 64, 128, 256, 320, 320], 'conv_op': 'torch.nn.modules.conv.Conv3d', 'kernel_sizes': [[3, 3, 3], [3, 3, 3], [3, 3, 3], [3, 3, 3], [3, 3, 3], [3, 3, 3]], 'strides': [[1, 1, 1], [2, 2, 2], [2, 2, 2], [2, 2, 2], [2, 2, 2], [2, 2, 2]], 'n_blocks_per_stage': [1, 3, 4, 6, 6, 6], 'n_conv_per_stage_decoder': [1, 1, 1, 1, 1], 'conv_bias': True, 'norm_op': 'torch.nn.modules.instancenorm.InstanceNorm3d', 'norm_op_kwargs': {'eps': 1e-05, 'affine': True}, 'dropout_op': None, 'dropout_op_kwargs': None, 'nonlin': 'torch.nn.LeakyReLU', 'nonlin_kwargs': {'inplace': True}}, '_kw_requires_import': ['conv_op', 'norm_op', 'dropout_op', 'nonlin']}, 'batch_dice': True}, '3d_cascade_fullres': {'inherits_from': '3d_fullres', 'previous_stage': '3d_lowres'}}, 'experiment_planner_used': 'nnUNetPlannerResEncL_torchres', 'label_manager': 'LabelManager', 'foreground_intensity_properties_per_channel': {'0': {'max': 1000.0, 'mean': -128.6943359375, 'median': 51.01091003417969, 'min': -1000.0, 'percentile_00_5': -996.0021362304688, 'percentile_99_5': 851.0109252929688, 'std': 398.4421691894531}}}",
46
+ "preprocessed_dataset_folder": "/projects/bodymaps/Pedro/data/nnUNet_preprocessed/Dataset307_radiologist_annotations_Merlin_Plus/nnUNetPlannerResEncL_torchres_isotropic_3d_fullres",
47
+ "preprocessed_dataset_folder_base": "/projects/bodymaps/Pedro/data/nnUNet_preprocessed/Dataset307_radiologist_annotations_Merlin_Plus",
48
+ "probabilistic_oversampling": "False",
49
+ "save_every": "50",
50
+ "torch_version": "2.6.0+cu124",
51
+ "was_initialized": "True",
52
+ "weight_decay": "3e-05"
53
+ }
Dataset307_radiologist_annotations_Merlin_Plus/nnUNetTrainer__nnUNetPlannerResEncL_torchres_isotropic__3d_fullres/fold_all/progress.png ADDED

Git LFS Details

  • SHA256: b7de5b318399784d6f393d0db44e547a5b95997d543910e2f096a2811adc0861
  • Pointer size: 131 Bytes
  • Size of remote file: 836 kB
Dataset307_radiologist_annotations_Merlin_Plus/nnUNetTrainer__nnUNetPlannerResEncL_torchres_isotropic__3d_fullres/fold_all/training_log_2026_2_21_12_15_45.txt ADDED
The diff for this file is too large to render. See raw diff
 
Dataset307_radiologist_annotations_Merlin_Plus/nnUNetTrainer__nnUNetPlannerResEncL_torchres_isotropic__3d_fullres/plans.json ADDED
@@ -0,0 +1,512 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "dataset_name": "Dataset307_radiologist_annotations_Merlin_Plus",
3
+ "plans_name": "nnUNetPlannerResEncL_torchres_isotropic",
4
+ "original_median_spacing_after_transp": [
5
+ 1.25,
6
+ 0.748046875,
7
+ 0.748046875
8
+ ],
9
+ "original_median_shape_after_transp": [
10
+ 409,
11
+ 512,
12
+ 512
13
+ ],
14
+ "image_reader_writer": "NibabelIOWithReorient",
15
+ "transpose_forward": [
16
+ 0,
17
+ 1,
18
+ 2
19
+ ],
20
+ "transpose_backward": [
21
+ 0,
22
+ 1,
23
+ 2
24
+ ],
25
+ "configurations": {
26
+ "2d": {
27
+ "data_identifier": "nnUNetPlannerResEncL_torchres_isotropic_2d",
28
+ "preprocessor_name": "DefaultPreprocessor",
29
+ "batch_size": 55,
30
+ "patch_size": [
31
+ 384,
32
+ 384
33
+ ],
34
+ "median_image_size_in_voxels": [
35
+ 383.0,
36
+ 383.0
37
+ ],
38
+ "spacing": [
39
+ 1.0,
40
+ 1.0
41
+ ],
42
+ "normalization_schemes": [
43
+ "CTNormalization"
44
+ ],
45
+ "use_mask_for_norm": [
46
+ false
47
+ ],
48
+ "resampling_fn_data": "resample_torch_fornnunet",
49
+ "resampling_fn_seg": "resample_torch_fornnunet",
50
+ "resampling_fn_data_kwargs": {
51
+ "is_seg": false,
52
+ "force_separate_z": false,
53
+ "memefficient_seg_resampling": false
54
+ },
55
+ "resampling_fn_seg_kwargs": {
56
+ "is_seg": true,
57
+ "force_separate_z": false,
58
+ "memefficient_seg_resampling": false
59
+ },
60
+ "resampling_fn_probabilities": "resample_torch_fornnunet",
61
+ "resampling_fn_probabilities_kwargs": {
62
+ "is_seg": false,
63
+ "force_separate_z": false,
64
+ "memefficient_seg_resampling": false
65
+ },
66
+ "architecture": {
67
+ "network_class_name": "dynamic_network_architectures.architectures.unet.ResidualEncoderUNet",
68
+ "arch_kwargs": {
69
+ "n_stages": 7,
70
+ "features_per_stage": [
71
+ 32,
72
+ 64,
73
+ 128,
74
+ 256,
75
+ 512,
76
+ 512,
77
+ 512
78
+ ],
79
+ "conv_op": "torch.nn.modules.conv.Conv2d",
80
+ "kernel_sizes": [
81
+ [
82
+ 3,
83
+ 3
84
+ ],
85
+ [
86
+ 3,
87
+ 3
88
+ ],
89
+ [
90
+ 3,
91
+ 3
92
+ ],
93
+ [
94
+ 3,
95
+ 3
96
+ ],
97
+ [
98
+ 3,
99
+ 3
100
+ ],
101
+ [
102
+ 3,
103
+ 3
104
+ ],
105
+ [
106
+ 3,
107
+ 3
108
+ ]
109
+ ],
110
+ "strides": [
111
+ [
112
+ 1,
113
+ 1
114
+ ],
115
+ [
116
+ 2,
117
+ 2
118
+ ],
119
+ [
120
+ 2,
121
+ 2
122
+ ],
123
+ [
124
+ 2,
125
+ 2
126
+ ],
127
+ [
128
+ 2,
129
+ 2
130
+ ],
131
+ [
132
+ 2,
133
+ 2
134
+ ],
135
+ [
136
+ 2,
137
+ 2
138
+ ]
139
+ ],
140
+ "n_blocks_per_stage": [
141
+ 1,
142
+ 3,
143
+ 4,
144
+ 6,
145
+ 6,
146
+ 6,
147
+ 6
148
+ ],
149
+ "n_conv_per_stage_decoder": [
150
+ 1,
151
+ 1,
152
+ 1,
153
+ 1,
154
+ 1,
155
+ 1
156
+ ],
157
+ "conv_bias": true,
158
+ "norm_op": "torch.nn.modules.instancenorm.InstanceNorm2d",
159
+ "norm_op_kwargs": {
160
+ "eps": 1e-05,
161
+ "affine": true
162
+ },
163
+ "dropout_op": null,
164
+ "dropout_op_kwargs": null,
165
+ "nonlin": "torch.nn.LeakyReLU",
166
+ "nonlin_kwargs": {
167
+ "inplace": true
168
+ }
169
+ },
170
+ "_kw_requires_import": [
171
+ "conv_op",
172
+ "norm_op",
173
+ "dropout_op",
174
+ "nonlin"
175
+ ]
176
+ },
177
+ "batch_dice": true
178
+ },
179
+ "3d_lowres": {
180
+ "data_identifier": "nnUNetPlannerResEncL_torchres_isotropic_3d_lowres",
181
+ "preprocessor_name": "DefaultPreprocessor",
182
+ "batch_size": 2,
183
+ "patch_size": [
184
+ 224,
185
+ 160,
186
+ 160
187
+ ],
188
+ "median_image_size_in_voxels": [
189
+ 319,
190
+ 261,
191
+ 261
192
+ ],
193
+ "spacing": [
194
+ 1.4685337134515648,
195
+ 1.4685337134515648,
196
+ 1.4685337134515648
197
+ ],
198
+ "normalization_schemes": [
199
+ "CTNormalization"
200
+ ],
201
+ "use_mask_for_norm": [
202
+ false
203
+ ],
204
+ "resampling_fn_data": "resample_torch_fornnunet",
205
+ "resampling_fn_seg": "resample_torch_fornnunet",
206
+ "resampling_fn_data_kwargs": {
207
+ "is_seg": false,
208
+ "force_separate_z": false,
209
+ "memefficient_seg_resampling": false
210
+ },
211
+ "resampling_fn_seg_kwargs": {
212
+ "is_seg": true,
213
+ "force_separate_z": false,
214
+ "memefficient_seg_resampling": false
215
+ },
216
+ "resampling_fn_probabilities": "resample_torch_fornnunet",
217
+ "resampling_fn_probabilities_kwargs": {
218
+ "is_seg": false,
219
+ "force_separate_z": false,
220
+ "memefficient_seg_resampling": false
221
+ },
222
+ "architecture": {
223
+ "network_class_name": "dynamic_network_architectures.architectures.unet.ResidualEncoderUNet",
224
+ "arch_kwargs": {
225
+ "n_stages": 6,
226
+ "features_per_stage": [
227
+ 32,
228
+ 64,
229
+ 128,
230
+ 256,
231
+ 320,
232
+ 320
233
+ ],
234
+ "conv_op": "torch.nn.modules.conv.Conv3d",
235
+ "kernel_sizes": [
236
+ [
237
+ 3,
238
+ 3,
239
+ 3
240
+ ],
241
+ [
242
+ 3,
243
+ 3,
244
+ 3
245
+ ],
246
+ [
247
+ 3,
248
+ 3,
249
+ 3
250
+ ],
251
+ [
252
+ 3,
253
+ 3,
254
+ 3
255
+ ],
256
+ [
257
+ 3,
258
+ 3,
259
+ 3
260
+ ],
261
+ [
262
+ 3,
263
+ 3,
264
+ 3
265
+ ]
266
+ ],
267
+ "strides": [
268
+ [
269
+ 1,
270
+ 1,
271
+ 1
272
+ ],
273
+ [
274
+ 2,
275
+ 2,
276
+ 2
277
+ ],
278
+ [
279
+ 2,
280
+ 2,
281
+ 2
282
+ ],
283
+ [
284
+ 2,
285
+ 2,
286
+ 2
287
+ ],
288
+ [
289
+ 2,
290
+ 2,
291
+ 2
292
+ ],
293
+ [
294
+ 2,
295
+ 2,
296
+ 2
297
+ ]
298
+ ],
299
+ "n_blocks_per_stage": [
300
+ 1,
301
+ 3,
302
+ 4,
303
+ 6,
304
+ 6,
305
+ 6
306
+ ],
307
+ "n_conv_per_stage_decoder": [
308
+ 1,
309
+ 1,
310
+ 1,
311
+ 1,
312
+ 1
313
+ ],
314
+ "conv_bias": true,
315
+ "norm_op": "torch.nn.modules.instancenorm.InstanceNorm3d",
316
+ "norm_op_kwargs": {
317
+ "eps": 1e-05,
318
+ "affine": true
319
+ },
320
+ "dropout_op": null,
321
+ "dropout_op_kwargs": null,
322
+ "nonlin": "torch.nn.LeakyReLU",
323
+ "nonlin_kwargs": {
324
+ "inplace": true
325
+ }
326
+ },
327
+ "_kw_requires_import": [
328
+ "conv_op",
329
+ "norm_op",
330
+ "dropout_op",
331
+ "nonlin"
332
+ ]
333
+ },
334
+ "batch_dice": false,
335
+ "next_stage": "3d_cascade_fullres"
336
+ },
337
+ "3d_fullres": {
338
+ "data_identifier": "nnUNetPlannerResEncL_torchres_isotropic_3d_fullres",
339
+ "preprocessor_name": "DefaultPreprocessor",
340
+ "batch_size": 2,
341
+ "patch_size": [
342
+ 224,
343
+ 160,
344
+ 160
345
+ ],
346
+ "median_image_size_in_voxels": [
347
+ 469.0,
348
+ 383.0,
349
+ 383.0
350
+ ],
351
+ "spacing": [
352
+ 1.0,
353
+ 1.0,
354
+ 1.0
355
+ ],
356
+ "normalization_schemes": [
357
+ "CTNormalization"
358
+ ],
359
+ "use_mask_for_norm": [
360
+ false
361
+ ],
362
+ "resampling_fn_data": "resample_torch_fornnunet",
363
+ "resampling_fn_seg": "resample_torch_fornnunet",
364
+ "resampling_fn_data_kwargs": {
365
+ "is_seg": false,
366
+ "force_separate_z": false,
367
+ "memefficient_seg_resampling": false
368
+ },
369
+ "resampling_fn_seg_kwargs": {
370
+ "is_seg": true,
371
+ "force_separate_z": false,
372
+ "memefficient_seg_resampling": false
373
+ },
374
+ "resampling_fn_probabilities": "resample_torch_fornnunet",
375
+ "resampling_fn_probabilities_kwargs": {
376
+ "is_seg": false,
377
+ "force_separate_z": false,
378
+ "memefficient_seg_resampling": false
379
+ },
380
+ "architecture": {
381
+ "network_class_name": "dynamic_network_architectures.architectures.unet.ResidualEncoderUNet",
382
+ "arch_kwargs": {
383
+ "n_stages": 6,
384
+ "features_per_stage": [
385
+ 32,
386
+ 64,
387
+ 128,
388
+ 256,
389
+ 320,
390
+ 320
391
+ ],
392
+ "conv_op": "torch.nn.modules.conv.Conv3d",
393
+ "kernel_sizes": [
394
+ [
395
+ 3,
396
+ 3,
397
+ 3
398
+ ],
399
+ [
400
+ 3,
401
+ 3,
402
+ 3
403
+ ],
404
+ [
405
+ 3,
406
+ 3,
407
+ 3
408
+ ],
409
+ [
410
+ 3,
411
+ 3,
412
+ 3
413
+ ],
414
+ [
415
+ 3,
416
+ 3,
417
+ 3
418
+ ],
419
+ [
420
+ 3,
421
+ 3,
422
+ 3
423
+ ]
424
+ ],
425
+ "strides": [
426
+ [
427
+ 1,
428
+ 1,
429
+ 1
430
+ ],
431
+ [
432
+ 2,
433
+ 2,
434
+ 2
435
+ ],
436
+ [
437
+ 2,
438
+ 2,
439
+ 2
440
+ ],
441
+ [
442
+ 2,
443
+ 2,
444
+ 2
445
+ ],
446
+ [
447
+ 2,
448
+ 2,
449
+ 2
450
+ ],
451
+ [
452
+ 2,
453
+ 2,
454
+ 2
455
+ ]
456
+ ],
457
+ "n_blocks_per_stage": [
458
+ 1,
459
+ 3,
460
+ 4,
461
+ 6,
462
+ 6,
463
+ 6
464
+ ],
465
+ "n_conv_per_stage_decoder": [
466
+ 1,
467
+ 1,
468
+ 1,
469
+ 1,
470
+ 1
471
+ ],
472
+ "conv_bias": true,
473
+ "norm_op": "torch.nn.modules.instancenorm.InstanceNorm3d",
474
+ "norm_op_kwargs": {
475
+ "eps": 1e-05,
476
+ "affine": true
477
+ },
478
+ "dropout_op": null,
479
+ "dropout_op_kwargs": null,
480
+ "nonlin": "torch.nn.LeakyReLU",
481
+ "nonlin_kwargs": {
482
+ "inplace": true
483
+ }
484
+ },
485
+ "_kw_requires_import": [
486
+ "conv_op",
487
+ "norm_op",
488
+ "dropout_op",
489
+ "nonlin"
490
+ ]
491
+ },
492
+ "batch_dice": true
493
+ },
494
+ "3d_cascade_fullres": {
495
+ "inherits_from": "3d_fullres",
496
+ "previous_stage": "3d_lowres"
497
+ }
498
+ },
499
+ "experiment_planner_used": "nnUNetPlannerResEncL_torchres",
500
+ "label_manager": "LabelManager",
501
+ "foreground_intensity_properties_per_channel": {
502
+ "0": {
503
+ "max": 1000.0,
504
+ "mean": -128.6943359375,
505
+ "median": 51.01091003417969,
506
+ "min": -1000.0,
507
+ "percentile_00_5": -996.0021362304688,
508
+ "percentile_99_5": 851.0109252929688,
509
+ "std": 398.4421691894531
510
+ }
511
+ }
512
+ }
PredictSubOrgansnUnet.py ADDED
@@ -0,0 +1,170 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ """
2
+ Use this to run the nnU-Net inference. You can run multiple instances of this code in parallel!
3
+ Just change the `part_id` and `gpu` arguments to split the task into multiple parts.
4
+ Example of splitting it in 4 parts:
5
+ python PredictSubOrgansnUnet.py --num_parts 4 --part_id 0 --gpu 0
6
+ python PredictSubOrgansnUnet.py --num_parts 4 --part_id 1 --gpu 1
7
+ python PredictSubOrgansnUnet.py --num_parts 4 --part_id 2 --gpu 2
8
+ python PredictSubOrgansnUnet.py --num_parts 4 --part_id 3 --gpu 3
9
+ """
10
+
11
+
12
+ import argparse
13
+ from nnunetv2.paths import nnUNet_results, nnUNet_raw
14
+ import torch
15
+ from batchgenerators.utilities.file_and_folder_operations import join
16
+ from nnunetv2.inference.predict_from_raw_data import nnUNetPredictor
17
+ import os
18
+ import pandas as pd
19
+
20
+ def split_files(files_input, files_output, num_parts, part_id):
21
+ """
22
+ Splits the files_input and files_output into num_parts and selects the partition for the given part_id.
23
+
24
+ Args:
25
+ files_input (list): List of input files.
26
+ files_output (list): List of output files.
27
+ num_parts (int): Number of partitions.
28
+ part_id (int): Partition ID to select (starting from 0).
29
+
30
+ Returns:
31
+ tuple: (files_input_partition, files_output_partition) for the given part_id.
32
+ """
33
+ assert len(files_input) == len(files_output), "files_input and files_output must have the same length"
34
+ assert 0 <= part_id < num_parts, "Invalid part_id. It must be in the range [0, num_parts-1]"
35
+
36
+ # Split the files based on num_parts
37
+ total_files = len(files_input)
38
+ files_per_part = (total_files + num_parts - 1) // num_parts # Ensure rounding up
39
+ start_idx = part_id * files_per_part
40
+ end_idx = min(start_idx + files_per_part, total_files)
41
+
42
+ # Return the partition
43
+ return files_input[start_idx:end_idx], files_output[start_idx:end_idx]
44
+
45
+ def parse_args():
46
+ parser = argparse.ArgumentParser(description="Run nnUNet prediction with custom arguments.")
47
+ parser.add_argument('--pth', type=str, default='/projects/bodymaps/Data/image_only/AbdomenAtlasPro/AbdomenAtlasPro/',
48
+ help="Path to input images folder")
49
+ parser.add_argument('--outdir', type=str, default='/projects/bodymaps/Pedro/data/JHH_liver_segments/',
50
+ help="Path to output folder for predictions")
51
+ parser.add_argument('--checkpoint', type=str,
52
+ default='./nnUNetOrgansAndSubSegments/',
53
+ help="Path to model checkpoint folder")
54
+ parser.add_argument('--num_parts', type=int, default=1,
55
+ help="Number of parts to split the task into")
56
+ parser.add_argument('--part_id', type=int, default=0,
57
+ help="ID of the current part (0-indexed)")
58
+ parser.add_argument('--gpu', type=int, default=0,
59
+ help="ID of the gpu to be used")
60
+ parser.add_argument('--workers', type=int, default=2,
61
+ help="Number of worker processes for preprocessing and segmentation export")
62
+ parser.add_argument('--BDMAP_format', action='store_true',
63
+ help="Enable BDMAP format")
64
+ parser.add_argument('--ids', default=None, help='Path to csv with a BDMAP ID column and the cases you want to inference')
65
+ parser.add_argument('--reset', action='store_true',
66
+ help="Overwrites old cases")
67
+ return parser.parse_args()
68
+
69
+
70
+ def filter_existing_outputs(files_input, files_output):
71
+ """
72
+ Filter out cases where the output file already exists.
73
+
74
+ Parameters:
75
+ files_input (list): A list of input file paths (or lists of file paths).
76
+ files_output (list): A list of output file paths.
77
+
78
+ Returns:
79
+ tuple: Two lists, (filtered_files_input, filtered_files_output), where
80
+ each pair corresponds to an output that does not already exist.
81
+ """
82
+ filtered_input = []
83
+ filtered_output = []
84
+
85
+ for inp, out in zip(files_input, files_output):
86
+ if not os.path.exists(out+'.nii.gz'):
87
+ filtered_input.append(inp)
88
+ filtered_output.append(out)
89
+ else:
90
+ print(f"Skipping already saved output: {out}")
91
+
92
+ return filtered_input, filtered_output
93
+
94
+ def main():
95
+ args = parse_args()
96
+
97
+ # Instantiate the nnUNetPredictor
98
+ predictor = nnUNetPredictor(
99
+ tile_step_size=0.5,
100
+ use_gaussian=True,
101
+ use_mirroring=True,
102
+ perform_everything_on_device=True,
103
+ device=torch.device('cuda', args.gpu),
104
+ verbose=False,
105
+ verbose_preprocessing=False,
106
+ allow_tqdm=True
107
+ )
108
+
109
+ if args.ids is not None:
110
+ ids = pd.read_csv(args.ids)
111
+ ids = sorted(ids['BDMAP ID'].to_list())
112
+ else:
113
+ ids = sorted(os.listdir(args.pth))
114
+
115
+ print(f'Number of loaded ids: {len(ids)}')
116
+
117
+
118
+ # Collect input and output files
119
+ if args.BDMAP_format:
120
+ print('Path:',args.pth)
121
+ print('Files in path:',len(ids))
122
+
123
+ files_input = [[os.path.join(args.pth, folder, 'ct.nii.gz')]
124
+ for folder in ids
125
+ if ('BDMAP' in folder)]
126
+ print('Input files before split:',len(files_input))
127
+ files_output = [os.path.join(args.outdir, folder)
128
+ for folder in ids
129
+ if ('BDMAP' in folder)]
130
+ files_input, files_output = filter_existing_outputs(files_input, files_output)
131
+ files_input, files_output = split_files(files_input, files_output, args.num_parts, args.part_id)
132
+ else:
133
+ if args.ids is None:
134
+ files_input = [[os.path.join(args.pth, file)] for file in ids if (file.endswith('.nii.gz') and file[0]!='.')]
135
+ files_output = [os.path.join(args.outdir, file.replace('.nii.gz','')) for file in ids if (file.endswith('.nii.gz') and file[0]!='.')]
136
+ else:
137
+ files_input = [[os.path.join(args.pth, file+'.nii.gz')] for file in ids]
138
+ files_output = [os.path.join(args.outdir, file) for file in ids]
139
+ files_input, files_output = filter_existing_outputs(files_input, files_output)
140
+
141
+ files_input, files_output = split_files(files_input, files_output, args.num_parts, args.part_id)
142
+
143
+
144
+ print('Input:', files_input[:10])
145
+ print('Output:', files_output[:10])
146
+ print(f'Cases to predict in part {args.part_id}: {len(files_input)}')
147
+
148
+
149
+ # Initialize the network architecture and load the checkpoint
150
+ predictor.initialize_from_trained_model_folder(
151
+ join(args.checkpoint),
152
+ use_folds=('all',),
153
+ checkpoint_name='checkpoint_final.pth',
154
+ )
155
+
156
+ # Run the prediction
157
+ predictor.predict_from_files(
158
+ files_input, files_output,
159
+ save_probabilities=False,
160
+ overwrite=args.reset,
161
+ num_processes_preprocessing=args.workers,
162
+ num_processes_segmentation_export=args.workers,
163
+ folder_with_segs_from_prev_stage=None,
164
+ num_parts=1,
165
+ part_id=0
166
+ )
167
+
168
+
169
+ if __name__ == '__main__':
170
+ main()
parallel_inference.sh ADDED
@@ -0,0 +1,90 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ #!/usr/bin/env bash
2
+ #
3
+ # Usage:
4
+ # bash parallel_inference.sh \
5
+ # --pth <path/to/dataset> \
6
+ # --outdir <path/to/output> \
7
+ # --checkpoint <path/to/model> \
8
+ # --gpus <gpu0,gpu1,...> \
9
+ # [--parts <part0,part1,...>] \
10
+ # [--num_parts <total_parts>] \
11
+ # [--ids <ids.csv>] [--BDMAP_format]
12
+ #
13
+ set -euo pipefail
14
+
15
+ # ---------- Parse CLI ---------------------------------------------------------
16
+ declare -a gpus=() parts=()
17
+ ids=""
18
+ bdmap_flag=false
19
+
20
+ while [[ $# -gt 0 ]]; do
21
+ case "$1" in
22
+ --pth) pth="$2"; shift 2 ;;
23
+ --outdir) outdir="$2"; shift 2 ;;
24
+ --checkpoint) checkpoint="$2"; shift 2 ;;
25
+ --gpus) IFS=',' read -r -a gpus <<< "$2"; shift 2 ;;
26
+ --parts) IFS=',' read -r -a parts <<< "$2"; shift 2 ;;
27
+ --num_parts) num_parts="$2"; shift 2 ;;
28
+ --ids) ids="$2"; shift 2 ;;
29
+ --BDMAP_format) bdmap_flag=true; shift 1 ;; # flag
30
+ *)
31
+ echo "Unknown option: $1" >&2
32
+ exit 1
33
+ ;;
34
+ esac
35
+ done
36
+
37
+ # ---------- Validate required args ------------------------------------------
38
+ if [[ -z "${pth:-}" || -z "${outdir:-}" || -z "${checkpoint:-}" || "${#gpus[@]}" -eq 0 ]]; then
39
+ cat >&2 <<EOF
40
+ Usage: bash $0 \
41
+ --pth <dataset> --outdir <output> --checkpoint <ckpt> \
42
+ --gpus gpu0,gpu1,... \
43
+ [--parts part0,part1,...] [--num_parts N] \
44
+ [--ids <ids.csv>] [--BDMAP_format]
45
+ EOF
46
+ exit 1
47
+ fi
48
+
49
+ mkdir -p logs
50
+
51
+ # ---------- Default parts & num_parts if missing ---------------------------
52
+ # default total parts = number of GPUs
53
+ if [[ -z "${num_parts:-}" ]]; then
54
+ num_parts=${#gpus[@]}
55
+ fi
56
+
57
+ # default parts = 0,1,2,...,num_parts-1
58
+ if [[ "${#parts[@]}" -eq 0 ]]; then
59
+ parts=()
60
+ for ((i=0; i<num_parts; i++)); do
61
+ parts+=("$i")
62
+ done
63
+ fi
64
+
65
+ # ---------- Launch per-part workers ------------------------------------------
66
+ gpu_count=${#gpus[@]}
67
+
68
+ for idx in "${!parts[@]}"; do
69
+ part_id="${parts[$idx]}"
70
+ # round-robin selection of GPU
71
+ gpu="${gpus[$(( idx % gpu_count ))]}"
72
+ logf="logs/part${part_id}_gpu${gpu}.log"
73
+ echo "Launching part $part_id / $num_parts on GPU $gpu → $logf"
74
+
75
+ cmd=(python PredictSubOrgansnUnet.py
76
+ --num_parts "$num_parts"
77
+ --part_id "$part_id"
78
+ --gpu "$gpu"
79
+ --pth "$pth"
80
+ --outdir "$outdir"
81
+ --checkpoint "$checkpoint")
82
+
83
+ [[ -n "$ids" ]] && cmd+=( --ids "$ids" )
84
+ [[ "$bdmap_flag" == true ]] && cmd+=( --BDMAP_format )
85
+
86
+ "${cmd[@]}" 2>&1 | tee -a "$logf" &
87
+ done
88
+
89
+ wait
90
+ echo "All jobs completed."
readme.md ADDED
@@ -0,0 +1,164 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ <p align="center">
2
+ <img src="https://huggingface.co/datasets/AbdomenAtlas/MerlinPlus/resolve/main/documents/tumor_examples.png" alt="Examples of tumor annotations in Merlin Plus" width="100%">
3
+ </p>
4
+
5
+ # Merlin-Cancer-Net
6
+
7
+ Merlin-Cancer-Net is an nnU-Net trained with tumor segmentation masks from the Merlin Plus dataset. It surpasses state-of-the-art public AI models and leading VLMs in the detection and segmentation of tumors in 9 organs: bladder, gallbladder, spleen, esophagus, stomach, duodenum, prostate, uterus, and adrenal glands.
8
+
9
+ **Merlin Plus** is a dataset with longitudinal metadata (patient IDs and scan dates) and per-voxel annotations for organs and 9 tumor types in the Merlin dataset (Stanford, 25,494 CT scans). Merlin Plus is part of a collaboration between the Merlin Project at Stanford and the R-Super Project at Johns Hopkins University. It is available at HuggingFace: https://huggingface.co/datasets/AbdomenAtlas/MerlinPlus. Merlin-Cancer-Net was trained with 423 masks from the Merlin Plus training dataset.
10
+
11
+
12
+ # Papers
13
+
14
+
15
+ <b>Scaling Artificial Intelligence for Multi-Tumor Early Detection with More Reports, Fewer Masks</b> <br/>
16
+ Pedro R. A. S. Bassi, Xinze Zhou, Wenxuan Li, Szymon Płotka, Jieneng Chen, Qi Chen, Zheren Zhu, Jakub Prządo, Ibrahim E. Hamacı, Sezgin Er, Yuhan Wang, Ashwin Kumar, Bjoern Menze, Jarosław B. Ćwikła, Yuyin Zhou, Akshay S. Chaudhari, Curtis P. Langlotz, Sergio Decherchi, Andrea Cavalli, Kang Wang, Yang Yang, Alan L. Yuille, Zongwei Zhou* <br/>
17
+ *Johns Hopkins University, Stanford University, University of Zurich, University of Warsaw, Italian Institute of Technology, UCSF* <br/>
18
+ Preprint <br/>
19
+ <a href='https://www.cs.jhu.edu/~zongwei/publication/bassi2025scaling.pdf'><img src='https://img.shields.io/badge/Paper-PDF-purple'></a>
20
+ <a href="https://github.com/MrGiovanni/R-Super">
21
+ <img src="https://img.shields.io/badge/GitHub-RSuper-black?logo=github">
22
+ </a>
23
+
24
+
25
+ **Merlin Plus: A Large-Scale, Multi-cancer, Image-Mask-Report Dataset**
26
+ Pedro R. A. S. Bassi†, Wenxuan Li†, Szymon Płotka†, Ruby Honjol, Jakub Prządo, Xinze Zhou, Kang Wang, Yang Yang, Malte Jensen, Akshay S. Chaudhari, Curtis P. Langlotz, Alan L. Yuille, and Zongwei Zhou.
27
+ *MICCAI 2026, LNCS 16895. Springer Nature Switzerland.*
28
+ <a href="https://papers.miccai.org/miccai-2026/paper/4063_paper.pdf"><img src="https://img.shields.io/badge/Paper-PDF-purple" alt="Paper PDF"></a>
29
+ <a href="https://huggingface.co/datasets/AbdomenAtlas/MerlinPlus">
30
+ <img src="https://img.shields.io/badge/🤗%20Hugging%20Face-Dataset-yellow">
31
+ </a>
32
+ <a href="https://stanford.redivis.com/datasets/bvj1-5hc2whwf6">
33
+ <img src="https://img.shields.io/badge/Stanford%20AIMI-Dataset-8C1515">
34
+ </a>
35
+ <a href="https://github.com/MrGiovanni/MerlinPlus">
36
+ <img src="https://img.shields.io/badge/GitHub-MerlinPlus-black?logo=github">
37
+ </a>
38
+
39
+
40
+ <b>Merlin: A Vision Language Foundation Model for 3D Computed Tomography</b> <br/>
41
+ Louis Blankemeier, Joseph Paul Cohen, Ashwin Kumar, Dave Van Veen, Syed Jamal Safdar Gardezi, Magdalini Paschali, Zhihong Chen, Jean-Benoit Delbrouck, Eduardo Reis, Cesar Truyts, Christian Bluethgen, Malte Engmann Kjeldskov Jensen, Sophie Ostmeier, Maya Varma, Jeya Maria Jose Valanarasu, Zhongnan Fang, Zepeng Huo, Zaid Nabulsi, Diego Ardila, Wei-Hung Weng, Edson Amaro Junior, Neera Ahuja, Jason Fries, Nigam H. Shah, Andrew Johnston, Robert D. Boutin, Andrew Wentland, Curtis P. Langlotz, Jason Hom, Sergios Gatidis, Akshay S. Chaudhari <br/>
42
+ *Stanford University* <br/>
43
+ *Nature, 2026.* <br/>
44
+ <a href='https://www.nature.com/articles/s41586-026-10181-8'><img src='https://img.shields.io/badge/Paper-PDF-purple'></a>
45
+ <a href="https://github.com/StanfordMIMI/Merlin">
46
+ <img src="https://img.shields.io/badge/GitHub-Merlin-black?logo=github">
47
+ </a>
48
+
49
+
50
+ <b>nnU-Net: a self-configuring method for deep learning-based biomedical image segmentation</b> <br/>
51
+ Fabian Isensee, Paul F. Jaeger, Simon A. A. Kohl, Jens Petersen, Klaus H. Maier-Hein <br/>
52
+ *German Cancer Research Center (DKFZ)* <br/>
53
+ *Nature Methods, 18(2), 203-211, 2021.* <br/>
54
+ <a href='https://www.nature.com/articles/s41592-020-01008-z'><img src='https://img.shields.io/badge/Paper-PDF-purple'></a>
55
+ <a href='https://github.com/MIC-DKFZ/nnUNet'><img src='https://img.shields.io/badge/GitHub-nnUNet-black'></a>
56
+
57
+ # Instructions
58
+
59
+ ### Installation
60
+
61
+ ```bash
62
+ conda create -n merlin_cancer_net python=3.12 -y
63
+ conda activate merlin_cancer_net
64
+ pip install torch torchvision torchaudio
65
+ pip install nnunetv2 pandas tqdm nibabel itk
66
+ ```
67
+
68
+ ### Data format
69
+
70
+ Put every CT in one folder, as `.nii.gz` files. You can use any file name.
71
+
72
+ ```
73
+ /path/to/dataset/
74
+ ├── case_1.nii.gz
75
+ ├── case_2.nii.gz
76
+ ...
77
+ ```
78
+
79
+ ### Inference
80
+
81
+ ```bash
82
+ bash parallel_inference.sh \
83
+ --pth /path/to/dataset/ \
84
+ --outdir /path/to/joint/outputs/ \
85
+ --checkpoint Dataset307_radiologist_annotations_Merlin_Plus/nnUNetTrainer__nnUNetPlannerResEncL_torchres_isotropic__3d_fullres/ \
86
+ --gpus 0
87
+ ```
88
+
89
+ Pass several GPUs as `--gpus 0,1,2,3` to run in parallel.
90
+
91
+ ### Split the joint labels
92
+
93
+ The nnU-Net produces joint labels (one file for all segmentation classes). To split these labels:
94
+
95
+ ```bash
96
+ python split_labels_multi_tumor.py \
97
+ --input_dir /path/to/joint/outputs/ \
98
+ --output_dir /path/to/outputs/
99
+ ```
100
+
101
+ Outputs one folder per case, with one mask per tumor and organ:
102
+
103
+ ```
104
+ /path/to/outputs/
105
+ ├── BDMAP_00000001/
106
+ | └── predictions/
107
+ | ├── spleen_lesion.nii.gz
108
+ | ├── stomach_lesion.nii.gz
109
+ | ├── liver.nii.gz
110
+ | └── ...
111
+ ```
112
+
113
+
114
+ # Citations
115
+
116
+ ```
117
+ @article{bassi2025scaling,
118
+ title={Scaling Artificial Intelligence for Multi-Tumor Early Detection with More Reports, Fewer Masks},
119
+ author={Bassi, Pedro R. A. S. and Zhou, Xinze and Li, Wenxuan and P{\l}otka, Szymon and Chen, Jieneng and Chen, Qi and Zhu, Zheren and Prz{\k{a}}do, Jakub and Hamamci, Ibrahim E. and Er, Sezgin and others},
120
+ journal={arXiv preprint arXiv:2510.14803},
121
+ year={2025}
122
+ }
123
+
124
+
125
+ @InProceedings{BasPed_Merlin_MICCAI2026,
126
+ author = { Bassi, Pedro R. A. S. AND Li, Wenxuan AND Płotka, Szymon AND Honjol, Ruby AND Prządo, Jakub AND Zhou, Xinze AND Wang, Kang AND Yang, Yang AND Jensen, Malte AND Chaudhari, Akshay S. AND Langlotz, Curtis P. AND Yuille, Alan L. AND Zhou, Zongwei},
127
+ title = { { Merlin Plus: A Large-Scale, Multi-cancer, Image-Mask-Report Dataset } },
128
+ booktitle = {Medical Image Computing and Computer Assisted Intervention -- MICCAI 2026},
129
+ year = {2026},
130
+ publisher = {Springer Nature Switzerland},
131
+ volume = {LNCS 16895},
132
+ month = {September},
133
+ page = {pending}
134
+ }
135
+
136
+
137
+ @article{blankemeier_kumar2026merlin,
138
+ author = {Blankemeier, Louis and Kumar, Ashwin and Cohen, Joseph Paul and Liu, Jiaming and Liu, Longchao and Van Veen, Dave and Gardezi, Syed Jamal Safdar and Yu, Hongkun and Paschali, Magdalini and Chen, Zhihong and Delbrouck, Jean-Benoit and Reis, Eduardo and Holland, Robbie and Truyts, Cesar and Bluethgen, Christian and Wu, Yufu and Lian, Long and Jensen, Malte Engmann Kjeldskov and Ostmeier, Sophie and Varma, Maya and Valanarasu, Jeya Maria Jose and Fang, Zhongnan and Huo, Zepeng and Nabulsi, Zaid and Ardila, Diego and Weng, Wei-Hung and Amaro Junior, Edson and Ahuja, Neera and Fries, Jason and Shah, Nigam H. and Zaharchuk, Greg and Willis, Marc and Yala, Adam and Johnston, Andrew and Boutin, Robert D. and Wentland, Andrew and Langlotz, Curtis P. and Hom, Jason and Gatidis, Sergios and Chaudhari, Akshay S.},
139
+ title = {Merlin: a computed tomography vision-language foundation model and dataset},
140
+ journal = {Nature},
141
+ year = {2026},
142
+ doi = {10.1038/s41586-026-10181-8},
143
+ url = {https://doi.org/10.1038/s41586-026-10181-8}
144
+ }
145
+
146
+ @article{isensee2021nnunet,
147
+ title={nnU-Net: a self-configuring method for deep learning-based biomedical image segmentation},
148
+ author={Isensee, Fabian and Jaeger, Paul F. and Kohl, Simon A. A. and Petersen, Jens and Maier-Hein, Klaus H.},
149
+ journal={Nature Methods},
150
+ volume={18},
151
+ number={2},
152
+ pages={203--211},
153
+ year={2021}
154
+ }
155
+ ```
156
+
157
+
158
+ ## Acknowledgement
159
+
160
+ This work was supported by the Lustgarten Foundation for Pancreatic Cancer Research, the Patrick J. McGovern Foundation Award, and the National Institutes of Health (NIH) under Award Number R01EB037669. Paper content is covered by patents pending.
161
+
162
+ © The Johns Hopkins University. This work is openly licensed via CC BY-NC-ND.
163
+ Creative Commons Attribution-NonCommercial-NoDerivatives 4.0
164
+ International Public License
split_labels_multi_tumor.py ADDED
@@ -0,0 +1,295 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ """
2
+ Usage example:
3
+ python split_labels.py --input_dir /projects/bodymaps/Pedro/data/JHH_liver_segments/ --output_dir /projects/bodymaps/Pedro/data/JHH_liver_segments_sep_labels/ --JHH --parts 3 --part 0
4
+ python split_labels.py --input_dir /projects/bodymaps/Data/UFO_OrganSubSegments_nnUNet/subSegments_output --output_dir /projects/bodymaps/Data/UCSF_pseudo_masks_subsegments --parts 3 --part 0
5
+
6
+ """
7
+
8
+ import os
9
+ import glob
10
+ import argparse
11
+ import numpy as np
12
+ import nibabel as nib
13
+ from concurrent.futures import ProcessPoolExecutor
14
+ import concurrent
15
+ from tqdm import tqdm
16
+
17
+
18
+ labels = ['background',
19
+ 'kidney_right',
20
+ 'kidney_left',
21
+ 'kidney_lesion',
22
+ 'kidney_lesion_kidney_right',
23
+ 'kidney_lesion_kidney_left',
24
+ 'pancreas',
25
+ 'pancreatic_lesion',
26
+ 'liver',
27
+ 'liver_lesion',
28
+ 'spleen',
29
+ 'spleen_lesion',
30
+ 'colon',
31
+ 'colon_lesion',
32
+ 'stomach',
33
+ 'stomach_lesion',
34
+ 'duodenum',
35
+ 'duodenum_lesion',
36
+ 'common_bile_duct',
37
+ 'intestine',
38
+ 'aorta',
39
+ 'postcava',
40
+ 'adrenal_gland_left',
41
+ 'adrenal_gland_right',
42
+ 'adrenal_lesion',
43
+ 'adrenal_lesion_adrenal_gland_left',
44
+ 'adrenal_lesion_adrenal_gland_right',
45
+ 'gall_bladder',
46
+ 'gallbladder_lesion',
47
+ 'gallbladder_lesion_gall_bladder',
48
+ 'bladder',
49
+ 'bladder_lesion',
50
+ 'esophagus',
51
+ 'esophagus_lesion',
52
+ 'hepatic_vessel',
53
+ 'portal_vein_and_splenic_vein',
54
+ 'lung_left',
55
+ 'lung_right',
56
+ 'lung_lesion',
57
+ 'lung_lesion_lung_left',
58
+ 'lung_lesion_lung_right',
59
+ 'prostate',
60
+ 'prostate_lesion',
61
+ 'uterus_lesion',
62
+ 'uterus_lesion_prostate',#this is a bug, our prostate annotation should actually be called prostate/uterus
63
+ 'femur_left',
64
+ 'femur_right',
65
+ 'superior_mesenteric_artery',
66
+ 'veins',
67
+ 'bone_lesion',
68
+ 'breast_lesion',]
69
+
70
+ labels = {name: i for i, name in enumerate(labels, 0)}
71
+
72
+
73
+ # Classes we want to output explicitly
74
+ #out_labels = [
75
+ # "liver",
76
+ # "pancreas",
77
+ # "kidney_right",
78
+ # "kidney_left",
79
+ # "liver_lesion",
80
+ # "kidney_lesion",
81
+ # "pancreatic_lesion",
82
+ # "pancreas_head",
83
+ # "pancreas_body",
84
+ # "pancreas_tail"
85
+ #]
86
+ # Add "liver_segment_1" through "liver_segment_8"
87
+ #for i in range(1, 9):
88
+ # out_labels.append("liver_segment_" + str(i))
89
+ out_labels = list(labels.keys())
90
+ #out_labels+=['liver','pancreas']
91
+
92
+ def parse_args():
93
+ parser = argparse.ArgumentParser(
94
+ description="Split segmentation labels into multiple .nii.gz files."
95
+ )
96
+ parser.add_argument(
97
+ "--input_dir",
98
+ type=str,
99
+ required=True,
100
+ help="Directory containing input .nii.gz files to process.",
101
+ )
102
+ parser.add_argument(
103
+ "--output_dir",
104
+ type=str,
105
+ required=True,
106
+ help="Directory where the splitted .nii.gz files will be saved.",
107
+ )
108
+ parser.add_argument(
109
+ "--num_workers",
110
+ type=int,
111
+ default=20,
112
+ help="Number of parallel processes to use for splitting.",
113
+ )
114
+ parser.add_argument(
115
+ "--restart",
116
+ action='store_true',
117
+ help="Overwrites files already saved.",
118
+ )
119
+
120
+ parser.add_argument(
121
+ "--parts",
122
+ type=int,
123
+ default=1,
124
+ help="Total number of parts to split the files into.",
125
+ )
126
+ parser.add_argument(
127
+ "--part",
128
+ type=int,
129
+ default=0,
130
+ help="Index of the part to process (0-indexed).",
131
+ )
132
+ parser.add_argument(
133
+ "--JHH",
134
+ action='store_true',
135
+ help="Processes only JHH files.",
136
+ )
137
+ return parser.parse_args()
138
+
139
+
140
+ def build_mapping():
141
+ """
142
+ Build the 'mapping' dictionary:
143
+ Each key is an output class name,
144
+ each value is a list of label IDs that should be merged into that class.
145
+ """
146
+
147
+ # Map from lesion label to the organ sub-parts it can overlap with
148
+ lesion_to_organs = {
149
+ 'kidney_lesion': ['kidney_right', 'kidney_left'],
150
+ 'adrenal_lesion': ['adrenal_gland_left', 'adrenal_gland_right'],
151
+ 'gallbladder_lesion': ['gall_bladder'],
152
+ 'lung_lesion': ['lung_left', 'lung_right'],
153
+ }
154
+
155
+ mapping = {}
156
+ for key in out_labels:
157
+ ids = []
158
+ for name, val in labels.items():
159
+ if name == key or name.startswith(key + "_"):
160
+ ids.append(val)
161
+
162
+ # Include overlap labels where this organ is the sub-segment
163
+ # e.g., kidney_lesion_kidney_right should be included for kidney_right
164
+ for lesion_name, organ_keys in lesion_to_organs.items():
165
+ if key in organ_keys:
166
+ overlap_label = lesion_name + "_" + key
167
+ if overlap_label in labels:
168
+ ids.append(labels[overlap_label])
169
+
170
+ # Special case: "pancreas" also includes "pancreatic"
171
+ if key == "pancreas":
172
+ for name, val in labels.items():
173
+ if name == "pancreatic_lesion" or name.startswith("pancreatic_lesion_"):
174
+ ids.append(val)
175
+
176
+ mapping[key] = sorted(set(ids))
177
+
178
+ return mapping
179
+
180
+
181
+
182
+ def split_segmentation(input_nifti_path, output_folder, classes_dict):
183
+ """
184
+ Reads a NIfTI segmentation, then saves separate .nii.gz masks
185
+ for each class (merging labels if necessary).
186
+ """
187
+ print('Splitting', input_nifti_path)
188
+ seg_img = nib.load(input_nifti_path)
189
+ # *** Fix: load as float, then cast to int16 (or directly from dataobj) ***
190
+ seg_data = seg_img.get_fdata().astype(np.int16)
191
+ # Or use: seg_data = np.asanyarray(seg_img.dataobj).astype(np.int16)
192
+
193
+ # Ensure output_folder exists
194
+ os.makedirs(output_folder, exist_ok=True)
195
+
196
+ # We'll name output files using:
197
+ # <base of input file>_<class_name>.nii.gz
198
+ base_name = os.path.basename(input_nifti_path)
199
+ if base_name.endswith(".nii.gz"):
200
+ base_name = base_name[:-7]
201
+ else:
202
+ base_name = os.path.splitext(base_name)[0]
203
+
204
+ # Optionally create a subfolder for each input file
205
+ this_out_folder = os.path.join(output_folder, base_name, 'predictions')
206
+ os.makedirs(this_out_folder, exist_ok=True)
207
+
208
+ for class_name, label_list in classes_dict.items():
209
+ # Create binary mask: 1 if label in label_list, else 0
210
+ mask = np.isin(seg_data, label_list).astype(np.uint8)
211
+
212
+ mask_nifti = nib.Nifti1Image(mask, seg_img.affine, seg_img.header)
213
+ out_file = os.path.join(this_out_folder, f"{class_name}.nii.gz")
214
+ nib.save(mask_nifti, out_file)
215
+
216
+ print(f"Finished splitting {input_nifti_path}!")
217
+
218
+ def process_file(input_path, output_dir, classes_dict):
219
+ """Wrapper to process a single file."""
220
+ split_segmentation(input_path, output_dir, classes_dict)
221
+
222
+ def main():
223
+ args = parse_args()
224
+ os.makedirs(args.output_dir, exist_ok=True)
225
+
226
+ # Build the label mapping once
227
+ classes_dict = build_mapping()
228
+ print(classes_dict)
229
+
230
+ # Gather all .nii.gz files in the input directory
231
+ nifti_files = glob.glob(os.path.join(args.input_dir, "*.nii.gz"))
232
+ if len(nifti_files) == 0:
233
+ print(f"No NIfTI files found in {args.input_dir}!")
234
+ return
235
+
236
+ if args.JHH:
237
+ nifti_files = [f for f in nifti_files if (('BDMAP_A' in f) or ('BDMAP_V' in f))]
238
+
239
+ # remove from the list all files aleady saved
240
+ if not args.restart:
241
+ new_nifti_files = []
242
+ for fpath in nifti_files:
243
+ base_name = os.path.basename(fpath)
244
+ if base_name.endswith(".nii.gz"):
245
+ base_name = base_name[:-7]
246
+ else:
247
+ base_name = os.path.splitext(base_name)[0]
248
+
249
+ # The same logic used in split_segmentation to build the output folder
250
+ this_out_folder = os.path.join(args.output_dir, base_name, 'predictions')
251
+
252
+ # Check if every class_name has a corresponding <class_name>.nii.gz
253
+ all_exist = True
254
+ for class_name in classes_dict.keys():
255
+ out_file = os.path.join(this_out_folder, f"{class_name}.nii.gz")
256
+ if not os.path.exists(out_file):
257
+ all_exist = False
258
+ break
259
+
260
+ # Keep this file if not all masks are present
261
+ if not all_exist:
262
+ new_nifti_files.append(fpath)
263
+
264
+ nifti_files = new_nifti_files
265
+
266
+ # Split the nifti_files list into parts and process only the selected part.
267
+ total_files = len(nifti_files)
268
+ if args.parts > 1:
269
+ files_per_part = (total_files + args.parts - 1) // args.parts # round up
270
+ start_idx = args.part * files_per_part
271
+ end_idx = min(start_idx + files_per_part, total_files)
272
+ nifti_files = nifti_files[start_idx:end_idx]
273
+ print(f"Processing part {args.part+1}/{args.parts}: files {start_idx} to {end_idx-1}")
274
+ else:
275
+ print("Processing all files (only one part).")
276
+ print(f"Total files to process: {len(nifti_files)}")
277
+
278
+ # Use parallel processing
279
+ with ProcessPoolExecutor(max_workers=args.num_workers) as executor:
280
+ futures = []
281
+ for fpath in nifti_files:
282
+ futures.append(
283
+ executor.submit(process_file, fpath, args.output_dir, classes_dict)
284
+ )
285
+ #for future in futures:
286
+ # future.result()
287
+ for future in tqdm(concurrent.futures.as_completed(futures), total=len(futures), desc="Processing files"):
288
+ future.result()
289
+
290
+ print(f"All files processed. Output saved to: {args.output_dir}")
291
+
292
+ if __name__ == "__main__":
293
+ main()
294
+
295
+ # version 2 - fixed label dict inversion, fixed substring matching in build_mapping (bladder/gall_bladder, prostate/uterus ambiguities)