lapix/CCAgT

Dataset Card for Images of Cervical Cells with AgNOR Stain Technique

Dataset Summary

The CCAgT (Images of Cervical Cells with AgNOR Stain Technique) dataset contains 9339 images (1600x1200 resolution where each pixel is 0.111µmX0.111µm) from 15 different slides stained using the AgNOR technique. Each image has at least one label. In total, this dataset has more than 63K instances of annotated object. The images are from the patients of the Gynecology and Colonoscopy Outpatient Clinic of the Polydoro Ernani de São Thiago University Hospital of the Universidade Federal de Santa Catarina (HU-UFSC).

Supported Tasks and Leaderboards

• image-segmentation: The dataset can be used to train a model for semantic segmentation or instance segmentation. Semantic segmentation consists in classifying each pixel of the image. Success on this task is typically measured by achieving high values of mean iou or f-score for pixels results. Instance segmentation consists of doing object detection first and then using a semantic segmentation model inside detected objects. For instances results, this task is typically measured by achieving high values of recall, precision and f-score.

• object-detection: The dataset can be used to train a model for object detection to detect the nuclei categories or the nucleolus organizer regions (NORs), which consists of locating instances of objects and then classifying each one. This task is typically measured by achieving a high values of recall, precision and f-score.

Languages

The class labels in the dataset are in English.

Dataset Structure

Data Instances

An example looks like the one below:

semantic segmentation (default configuration)

{
  'image': <PIL.JpegImagePlugin.JpegImageFile image mode=RGB size=1200x1600 at 0x276021C5EB8>,
  'annotation': <PIL.PngImagePlugin.PngImageFile image mode=L size=1200x1600 at 0x385021C5ED7>
}

object detection

{
  'image': <PIL.JpegImagePlugin.JpegImageFile image mode=RGB size=1200x1600 at 0x276021C5EB8>,
  'objects': {
    'bbox': [
      [36, 7, 13, 32],
      [50, 7, 12, 32]
    ], 
    'label': [1, 5]
  }

instance segmentation

{
  'image': <PIL.JpegImagePlugin.JpegImageFile image mode=RGB size=1200x1600 at 0x276021C5EB8>,
  'objects': {
    'bbox': [
      [13.3, 7.5, 47.6, 38.3],
      [10.2, 7.5, 50.7, 38.3]
    ], 
    'segment': [
      [[36.2, 7.5, 13.3, 32.1, 52.1, 40.6, 60.9, 45.8, 50.1, 40, 40, 33.2, 35.2]],
      [[10.2, 7.5, 10.3, 32.1, 52.1, 40.6, 60.9, 45.8, 50.1, 40, 40, 33.2, 35.2]],
    ],
    'label': [1, 5]
  }

Data Fields

The data annotations have the following fields:

semantic segmentation (default configuration)

• image: A PIL.Image.Image object containing the image. Note that when accessing the image column: dataset[0]["image"] the image file is automatically decoded. Decoding of a large number of image files might take a significant amount of time. Thus it is important to first query the sample index before the "image" column, i.e. dataset[0]["image"] should always be preferred over dataset["image"][0].
• annotation: A PIL.Image.Image object containing the annotation mask. The mask has a single channel and the following pixel values are possible: BACKGROUND (0), NUCLEUS (1), CLUSTER (2), SATELLITE (3), NUCLEUS_OUT_OF_FOCUS (4), OVERLAPPED_NUCLEI (5), NON_VIABLE_NUCLEUS (6) and LEUKOCYTE_NUCLEUS (7).

object detection

• image: A PIL.Image.Image object containing the image. Note that when accessing the image column: dataset[0]["image"] the image file is automatically decoded. Decoding of a large number of image files might take a significant amount of time. Thus it is important to first query the sample index before the "image" column, i.e. dataset[0]["image"] should always be preferred over dataset["image"][0].
• objects: a dictionary containing bounding boxes and labels of the cell objects
  • bbox: a list of bounding boxes (in the coco format) corresponding to the objects present on the image
  • label: a list of integers representing the category (7 categories to describe the objects in total; two to differentiate nucleolus organizer regions), with the possible values including NUCLEUS (0), CLUSTER (1), SATELLITE (2), NUCLEUS_OUT_OF_FOCUS (3), OVERLAPPED_NUCLEI (4), NON_VIABLE_NUCLEUS (5) and LEUKOCYTE_NUCLEUS (6).

instance segmentation

• image: A PIL.Image.Image object containing the image. Note that when accessing the image column: dataset[0]["image"] the image file is automatically decoded. Decoding of a large number of image files might take a significant amount of time. Thus it is important to first query the sample index before the "image" column, i.e. dataset[0]["image"] should always be preferred over dataset["image"][0].
• objects: a dictionary containing bounding boxes and labels of the cell objects
  • bbox: a list of bounding boxes (in the coco format) corresponding to the objects present on the image
  • segment: a list of segments in format of [polygon_0, ..., polygon_n], where each polygon is [x0, y0, ..., xn, yn].
  • label: a list of integers representing the category (7 categories to describe the objects in total; two to differentiate nucleolus organizer regions), with the possible values including NUCLEUS (0), CLUSTER (1), SATELLITE (2), NUCLEUS_OUT_OF_FOCUS (3), OVERLAPPED_NUCLEI (4), NON_VIABLE_NUCLEUS (5) and LEUKOCYTE_NUCLEUS (6).

Data Splits

The data is split randomly using the fixed seed into training, test and validation set. The training data contains 70% of the images and the testing and the validation data contain 15% of the images each. In total, the training set contains 6533 images and the testing and the validation set 1403 images each.

Click here to see additional statistics:

Slide id	Diagnostics	images	annotations	NUCLEUS	CLUSTER	SATELLITE	NUCLEUS_OUT_OF_FOCUS	OVERLAPPED_NUCLEI	NON_VIABLE_NUCLEUS	LEUKOCYTE_NUCLEUS
A	CIN 3	1311	3164	763	1038	922	381	46	14	0
B	SCC	561	911	224	307	112	132	5	1	130
C	AC	385	11420	2420	3584	1112	1692	228	477	1907
D	CIN 3	2125	1258	233	337	107	149	12	8	412
E	CIN 3	506	11131	2611	6249	1648	476	113	34	0
F	CIN 1	318	3365	954	1406	204	354	51	326	70
G	CIN 2	249	2759	691	1279	336	268	49	51	85
H	CIN 2	650	5216	993	983	425	2562	38	214	1
I	No lesion	309	474	56	55	19	170	2	23	149
J	CIN 1	261	1786	355	304	174	743	18	33	159
K	No lesion	1503	13102	2464	6669	638	620	670	138	1903
L	CIN 2	396	3289	842	796	387	1209	27	23	5
M	CIN 2	254	1500	357	752	99	245	16	12	19
N	CIN 3	248	911	258	402	67	136	10	6	32
O	AC	262	2904	792	1549	228	133	88	52	62
Total	-	9339	63190	14013	25710	6478	9270	1373	1412	4934

Lesion types:

• Cervical intraepithelial neoplasia 1 - CIN 1
• Cervical intraepithelial neoplasia 2 - CIN 2
• Cervical intraepithelial neoplasia 3 - CIN 3
• Squamous cell carcinoma - SCC
• Adenocarcinoma - AC
• No lesion

Dataset Creation

Curation Rationale

CCAgT was built to provide a dataset for machines to learn how to identify nucleus and nucleolus organizer regions (NORs).

Source Data

Initial Data Collection and Normalization

The images are collected as patches/tiles of whole slide images (WSIs) from cervical samples stained with AgNOR technique to allow the detection of nucleolus organizer regions (NORs). NORs are DNA loops containing genes responsible for the transcription of ribosomal RNA located in the cell nucleolus. They contain a set of argyrophilic proteins, selectively stained by silver nitrate, which can be identified as black dots located throughout the nucleoli area and called AgNORs.

Who are the source language producers?

The dataset was built using images from examinations (a gynecological exam, colposcopy and biopsy) of 15 women patients who were treated at the Gynecology and Colposcopy Outpatient Clinic of the University Hospital Professor Polydoro Ernani de São Thiago of Federal University of Santa Catarina (HU-UFSC) and had 6 different diagnoses in their oncological exams. The samples were collected by the members of the Clinical Analyses Department: Ane Francyne Costa, Fabiana Botelho De Miranda Onofre, and Alexandre Sherlley Casimiro Onofre.

Annotations

Annotation process

The instances were annotated using the labelbox tool. The satellite category was labeled as a single dot, and the other categories were labeled as polygons. After the annotation process, all annotations were reviewed.

Who are the annotators?

Members of the Clinical Analyses Department and the Image Processing and Computer Graphics Lab. — LAPiX from Universidade Federal de Santa Catarina (UFSC).

• Tainee Bottamedi
• Vinícius Sanches
• João H. Telles de Carvalho
• Ricardo Thisted

Personal and Sensitive Information

This research was approved by the UFSC Research Ethics Committee (CEPSH), protocol number 57423616.3.0000.0121. All involved patients were informed about the study's objectives, and those who agreed to participate signed an informed consent form.

Considerations for Using the Data

Social Impact of Dataset

This dataset's purpose is to help spread the AgNOR as a support method for cancer diagnosis since this method is not standardized among pathologists.

Discussion of Biases

[More Information Needed]

Other Known Limitations

Satellite annotation is not as accurate for pixel-level representation due to single-point annotations.

Additional Information

Dataset Curators

Members of the Clinical Analyses Department from Universidade Federal de Santa Catarina (UFSC) collected the dataset samples: Ane Francyne Costa, Fabiana Botelho De Miranda Onofre, and Alexandre Sherlley Casimiro Onofre.

Licensing Information

The files associated with this dataset are licensed under an Attribution-NonCommercial 3.0 Unported license.

Users are free to adapt, copy or redistribute the material as long as they attribute it appropriately and do not use it for commercial purposes.

Citation Information

% Dataset oficial page
@misc{CCAgTDataset,
  doi = {10.17632/WG4BPM33HJ.2},
  url = {https://data.mendeley.com/datasets/wg4bpm33hj/2},
  author =  {Jo{\~{a}}o Gustavo Atkinson Amorim and Andr{\'{e}} Vict{\'{o}}ria Matias and Tainee Bottamedi and Vin{\'{i}}us Sanches and Ane Francyne Costa and Fabiana Botelho De Miranda Onofre and Alexandre Sherlley Casimiro Onofre and Aldo von Wangenheim},
  title = {CCAgT: Images of Cervical Cells with AgNOR Stain Technique},
  publisher = {Mendeley},
  year = {2022},
  copyright = {Attribution-NonCommercial 3.0 Unported}
}


% Dataset second version
% pre-print:
@article{AtkinsonAmorim2022,
  doi = {10.2139/ssrn.4126881},
  url = {https://doi.org/10.2139/ssrn.4126881},
  year = {2022},
  publisher = {Elsevier {BV}},
  author = {Jo{\~{a}}o Gustavo Atkinson Amorim and Andr{\'{e}} Vict{\'{o}}ria Matias and Allan Cerentini and Fabiana Botelho de Miranda Onofre and Alexandre Sherlley Casimiro Onofre and Aldo von Wangenheim},
  title = {Semantic Segmentation for the Detection of Very Small Objects on Cervical Cell Samples Stained with the {AgNOR} Technique},
  journal = {{SSRN} Electronic Journal}
}


% Dataset first version
% Link: https://arquivos.ufsc.br/d/373be2177a33426a9e6c/
% Paper:
@inproceedings{AtkinsonSegmentationAgNORCBMS2020,  
    author={Jo{\~{a}}o Gustavo Atkinson Amorim and Luiz Antonio Buschetto Macarini and Andr{\'{e}} Vict{\'{o}}ria Matias and Allan Cerentini and Fabiana Botelho De Miranda Onofre and Alexandre Sherlley Casimiro Onofre and Aldo von Wangenheim},
    booktitle={2020 IEEE 33rd International Symposium on Computer-Based Medical Systems (CBMS)},
    title={A Novel Approach on Segmentation of AgNOR-Stained Cytology Images Using Deep Learning},
    year={2020},
    pages={552-557}, 
    doi={10.1109/CBMS49503.2020.00110},
    url={https://doi.org/10.1109/CBMS49503.2020.00110}
}

Contributions

Thanks to @johnnv1 for adding this dataset.