Datasets:
image imagewidth (px) 256 256 | inst_map imagewidth (px) 256 256 | class_map imagewidth (px) 256 256 | patch_id int32 0 4.98k | patch_info stringlengths 11 15 | source stringclasses 5
values | count_neutrophil int32 0 50 | count_epithelial int32 0 253 | count_lymphocyte int32 0 450 | count_plasma int32 0 90 | count_eosinophil int32 0 25 | count_connective int32 0 104 |
|---|---|---|---|---|---|---|---|---|---|---|---|
0 | consep_1-0000 | consep | 0 | 117 | 0 | 0 | 0 | 0 | |||
1 | consep_1-0001 | consep | 0 | 95 | 1 | 0 | 0 | 8 | |||
2 | consep_1-0002 | consep | 0 | 172 | 3 | 0 | 0 | 2 | |||
3 | consep_1-0003 | consep | 0 | 56 | 0 | 0 | 0 | 10 | |||
4 | consep_10-0000 | consep | 0 | 169 | 7 | 0 | 0 | 0 | |||
5 | consep_10-0001 | consep | 0 | 181 | 4 | 0 | 0 | 8 | |||
6 | consep_10-0002 | consep | 0 | 118 | 5 | 2 | 0 | 6 | |||
7 | consep_10-0003 | consep | 0 | 152 | 8 | 0 | 0 | 7 | |||
8 | consep_11-0000 | consep | 0 | 45 | 49 | 7 | 3 | 26 | |||
9 | consep_11-0001 | consep | 0 | 60 | 37 | 14 | 0 | 17 | |||
10 | consep_11-0002 | consep | 0 | 67 | 31 | 4 | 0 | 18 | |||
11 | consep_11-0003 | consep | 0 | 53 | 32 | 9 | 1 | 27 | |||
12 | consep_12-0000 | consep | 0 | 0 | 2 | 2 | 0 | 2 | |||
13 | consep_12-0001 | consep | 0 | 0 | 7 | 0 | 0 | 11 | |||
14 | consep_12-0002 | consep | 0 | 0 | 8 | 0 | 0 | 35 | |||
15 | consep_12-0003 | consep | 0 | 0 | 5 | 2 | 0 | 26 | |||
16 | consep_13-0000 | consep | 0 | 55 | 11 | 24 | 2 | 28 | |||
17 | consep_13-0001 | consep | 0 | 69 | 19 | 30 | 1 | 24 | |||
18 | consep_13-0002 | consep | 0 | 56 | 10 | 33 | 1 | 27 | |||
19 | consep_13-0003 | consep | 0 | 46 | 17 | 53 | 4 | 44 | |||
20 | consep_14-0000 | consep | 0 | 58 | 1 | 0 | 0 | 0 | |||
21 | consep_14-0001 | consep | 0 | 61 | 1 | 0 | 0 | 6 | |||
22 | consep_14-0002 | consep | 0 | 46 | 4 | 0 | 0 | 11 | |||
23 | consep_14-0003 | consep | 0 | 59 | 0 | 0 | 0 | 6 | |||
24 | consep_15-0000 | consep | 1 | 0 | 17 | 3 | 6 | 20 | |||
25 | consep_15-0001 | consep | 0 | 0 | 10 | 2 | 3 | 29 | |||
26 | consep_15-0002 | consep | 0 | 0 | 7 | 4 | 3 | 31 | |||
27 | consep_15-0003 | consep | 1 | 0 | 7 | 1 | 3 | 20 | |||
28 | consep_16-0000 | consep | 1 | 0 | 6 | 0 | 0 | 24 | |||
29 | consep_16-0001 | consep | 5 | 0 | 7 | 2 | 0 | 16 | |||
30 | consep_16-0002 | consep | 4 | 0 | 5 | 0 | 0 | 6 | |||
31 | consep_16-0003 | consep | 2 | 0 | 3 | 0 | 0 | 15 | |||
32 | consep_2-0000 | consep | 0 | 0 | 16 | 19 | 0 | 45 | |||
33 | consep_2-0001 | consep | 0 | 0 | 117 | 3 | 0 | 67 | |||
34 | consep_2-0002 | consep | 0 | 0 | 61 | 10 | 3 | 81 | |||
35 | consep_2-0003 | consep | 0 | 0 | 161 | 3 | 0 | 68 | |||
36 | consep_3-0000 | consep | 0 | 0 | 15 | 0 | 0 | 21 | |||
37 | consep_3-0001 | consep | 0 | 0 | 10 | 0 | 0 | 20 | |||
38 | consep_3-0002 | consep | 0 | 0 | 21 | 2 | 0 | 27 | |||
39 | consep_3-0003 | consep | 1 | 0 | 56 | 8 | 1 | 23 | |||
40 | consep_4-0000 | consep | 0 | 0 | 1 | 0 | 0 | 52 | |||
41 | consep_4-0001 | consep | 0 | 0 | 2 | 0 | 0 | 33 | |||
42 | consep_4-0002 | consep | 0 | 0 | 3 | 0 | 0 | 48 | |||
43 | consep_4-0003 | consep | 0 | 0 | 3 | 0 | 0 | 32 | |||
44 | consep_5-0000 | consep | 0 | 88 | 2 | 0 | 0 | 1 | |||
45 | consep_5-0001 | consep | 0 | 91 | 5 | 1 | 0 | 7 | |||
46 | consep_5-0002 | consep | 0 | 91 | 20 | 4 | 0 | 13 | |||
47 | consep_5-0003 | consep | 0 | 112 | 2 | 0 | 0 | 3 | |||
48 | consep_6-0000 | consep | 0 | 39 | 9 | 2 | 0 | 18 | |||
49 | consep_6-0001 | consep | 0 | 28 | 14 | 1 | 0 | 23 | |||
50 | consep_6-0002 | consep | 0 | 30 | 37 | 9 | 0 | 23 | |||
51 | consep_6-0003 | consep | 0 | 41 | 9 | 2 | 0 | 10 | |||
52 | consep_7-0000 | consep | 8 | 0 | 3 | 0 | 0 | 1 | |||
53 | consep_7-0001 | consep | 0 | 0 | 1 | 0 | 0 | 6 | |||
54 | consep_7-0002 | consep | 0 | 0 | 0 | 0 | 0 | 5 | |||
55 | consep_7-0003 | consep | 0 | 0 | 0 | 0 | 0 | 7 | |||
56 | consep_8-0000 | consep | 0 | 88 | 48 | 8 | 1 | 22 | |||
57 | consep_8-0001 | consep | 0 | 49 | 114 | 9 | 10 | 17 | |||
58 | consep_8-0002 | consep | 0 | 91 | 27 | 3 | 4 | 21 | |||
59 | consep_8-0003 | consep | 0 | 52 | 22 | 2 | 0 | 13 | |||
60 | consep_9-0000 | consep | 4 | 0 | 4 | 0 | 0 | 6 | |||
61 | consep_9-0001 | consep | 0 | 0 | 0 | 0 | 0 | 14 | |||
62 | consep_9-0002 | consep | 0 | 0 | 2 | 0 | 0 | 0 | |||
63 | consep_9-0003 | consep | 0 | 0 | 1 | 0 | 0 | 16 | |||
64 | crag_1-0000 | crag | 0 | 0 | 0 | 0 | 0 | 33 | |||
65 | crag_1-0001 | crag | 0 | 0 | 1 | 0 | 0 | 9 | |||
66 | crag_1-0002 | crag | 0 | 0 | 0 | 0 | 0 | 11 | |||
67 | crag_1-0003 | crag | 0 | 0 | 0 | 0 | 0 | 10 | |||
68 | crag_1-0004 | crag | 0 | 0 | 0 | 0 | 0 | 5 | |||
69 | crag_1-0005 | crag | 0 | 0 | 1 | 0 | 0 | 35 | |||
70 | crag_1-0006 | crag | 0 | 0 | 4 | 0 | 0 | 26 | |||
71 | crag_1-0007 | crag | 0 | 0 | 0 | 0 | 0 | 16 | |||
72 | crag_1-0008 | crag | 0 | 0 | 1 | 0 | 0 | 14 | |||
73 | crag_1-0009 | crag | 0 | 0 | 2 | 0 | 0 | 18 | |||
74 | crag_1-0010 | crag | 0 | 0 | 1 | 0 | 0 | 32 | |||
75 | crag_1-0011 | crag | 0 | 0 | 3 | 0 | 0 | 34 | |||
76 | crag_1-0012 | crag | 0 | 0 | 1 | 0 | 0 | 18 | |||
77 | crag_1-0013 | crag | 0 | 0 | 0 | 0 | 1 | 12 | |||
78 | crag_1-0014 | crag | 0 | 0 | 0 | 0 | 0 | 8 | |||
79 | crag_1-0015 | crag | 0 | 0 | 1 | 0 | 0 | 23 | |||
80 | crag_1-0016 | crag | 0 | 4 | 1 | 0 | 0 | 35 | |||
81 | crag_1-0017 | crag | 0 | 3 | 0 | 0 | 0 | 21 | |||
82 | crag_1-0018 | crag | 0 | 0 | 1 | 0 | 0 | 3 | |||
83 | crag_1-0019 | crag | 0 | 0 | 0 | 0 | 0 | 1 | |||
84 | crag_1-0020 | crag | 0 | 14 | 1 | 1 | 0 | 37 | |||
85 | crag_1-0021 | crag | 0 | 39 | 1 | 0 | 0 | 19 | |||
86 | crag_1-0022 | crag | 0 | 28 | 1 | 0 | 0 | 17 | |||
87 | crag_1-0023 | crag | 0 | 16 | 2 | 0 | 0 | 9 | |||
88 | crag_1-0024 | crag | 0 | 0 | 0 | 0 | 0 | 0 | |||
89 | crag_1-0025 | crag | 0 | 0 | 5 | 1 | 0 | 42 | |||
90 | crag_1-0026 | crag | 0 | 0 | 2 | 0 | 0 | 29 | |||
91 | crag_1-0027 | crag | 0 | 2 | 0 | 0 | 0 | 23 | |||
92 | crag_1-0028 | crag | 0 | 0 | 0 | 0 | 0 | 11 | |||
93 | crag_1-0029 | crag | 0 | 0 | 0 | 0 | 0 | 0 | |||
94 | crag_1-0030 | crag | 0 | 1 | 0 | 0 | 0 | 11 | |||
95 | crag_1-0031 | crag | 0 | 20 | 1 | 0 | 0 | 8 | |||
96 | crag_1-0032 | crag | 0 | 2 | 0 | 0 | 0 | 20 | |||
97 | crag_1-0033 | crag | 0 | 18 | 0 | 0 | 0 | 3 | |||
98 | crag_1-0034 | crag | 0 | 0 | 0 | 0 | 0 | 1 | |||
99 | crag_1-0035 | crag | 0 | 0 | 0 | 0 | 0 | 0 |
CoNIC 2022 (Colon Nuclei Identification and Counting)
Public training release of the CoNIC 2022 Challenge: 4,981 non-overlapping 256x256 H&E histopathology patches (colon tissue, 20x, ~0.5 um/pixel) with nucleus instance segmentation, classification and counting annotations.
Faithful-naming note. This mirror contains the public CoNIC training split only (4,981 patches). The challenge test set (~103k nuclei) is permanently held out and is not part of this dataset. The patches are a re-tiled view of the Lizard dataset (whole-image
.matannotations cut into non-overlapping 256x256 tiles).
Composition
- 4,981 RGB patches of 256x256 px, 431,913 annotated nuclei
- 6 nucleus classes + background
- Single
trainsplit (the only public release; no official fold split)
Lizard source breakdown (source column)
| source | patches | origin dataset |
|---|---|---|
crag |
2,304 | CRAG |
dpath |
1,799 | DigestPath |
glas |
702 | GlaS |
pannuke |
112 | PanNuke |
consep |
64 | CoNSeP |
Schema
| Field | Type | Description |
|---|---|---|
image |
PIL RGB 256x256 | H&E patch, uint8 |
inst_map |
PIL grayscale uint16 256x256 | Per-patch unique nucleus instance ID (0 = background) |
class_map |
PIL grayscale uint8 256x256 | Semantic class per pixel (see table below) |
patch_id |
int | Row index 0..4980 (matches the original .npy ordering) |
patch_info |
str | Lizard source image token, e.g. consep_1-0000 |
source |
str | Lizard source dataset (crag/dpath/glas/pannuke/consep) |
count_* |
int (x6) | Per-class nucleus count in the central 224x224 region |
Semantic class encoding (class_map)
| Value | Class |
|---|---|
| 0 | Background |
| 1 | Neutrophil |
| 2 | Epithelial |
| 3 | Lymphocyte |
| 4 | Plasma |
| 5 | Eosinophil |
| 6 | Connective |
count_* columns (count_neutrophil ... count_connective) follow the same
1..6 ordering and are taken from the challenge counts.csv (counted within the
central 224x224 region only, per the official protocol).
Cross-dataset overlap (leakage warning)
CoNIC is a re-tiled subset of Lizard, which is itself assembled from CRAG,
DigestPath, GlaS, PanNuke and CoNSeP. In particular it shares source images
with Angelou0516/PanNuke
(112 patches), as well as GlaS (702), CRAG (2,304), DigestPath (1,799) and
CoNSeP (64). Use the source / patch_info columns to perform source-image
level deduplication before co-benchmarking against any of these datasets.
License
CC BY-NC-SA 4.0 (research / non-commercial use). Same license as the original CoNIC / Lizard release.
Citation
@article{graham2024conic,
title={CoNIC Challenge: Pushing the frontiers of nuclear detection, segmentation, classification and counting},
author={Graham, Simon and Vu, Quoc Dang and Jahanifar, Mostafa and Weigert, Martin and Schmidt, Uwe and Zhang, Wenhua and others},
journal={Medical Image Analysis},
volume={92},
pages={103047},
year={2024},
publisher={Elsevier}
}
@article{graham2021conic,
title={CoNIC: Colon Nuclei Identification and Counting Challenge 2022},
author={Graham, Simon and Jahanifar, Mostafa and Vu, Quoc Dang and Hadjigeorghiou, Giorgos and Leech, Thomas and Snead, David and Raza, Shan E Ahmed and Minhas, Fayyaz and Rajpoot, Nasir},
journal={arXiv preprint arXiv:2111.14485},
year={2021}
}
@inproceedings{graham2021lizard,
title={Lizard: A large-scale dataset for colonic nuclear instance segmentation and classification},
author={Graham, Simon and Jahanifar, Mostafa and Azam, Ayesha and Nimir, Mohammed and Tsang, Yee-Wah and Dodd, Katherine and Hero, Emily and Sahota, Harvir and Tank, Atisha and Benes, Ksenija and others},
booktitle={Proceedings of the IEEE/CVF International Conference on Computer Vision},
pages={684--693},
year={2021}
}
- Downloads last month
- 26