Datasets:

bumbledeep
/

aptos

dataset_info:
  features:
    - name: image
      dtype: image
    - name: label_code
      dtype: int64
    - name: label
      dtype: string
  splits:
    - name: train
      num_bytes: 249970283.654
      num_examples: 3662
  download_size: 249981721
  dataset_size: 249970283.654
configs:
  - config_name: default
    data_files:
      - split: train
        path: data/train-*
license: mit

Dataset Card for Dataset Name

This dataset card aims to be a base template for new datasets. It has been generated using this raw template.

Dataset Details

Dataset Description

Asia Pacific Tele-Ophthalmology Society (APTOS) dataset. The images consist of retina scan images to detect diabetic retinopathy. The original dataset is available at APTOS 2019 Blindness Detection. These images are resized into 224x224 pixels so that they can be readily used with many pre-trained deep learning models.

Funded by [optional]: Asia Pacific Tele-Ophthalmology Society (APTOS).
Shared by: Sovit Ranjan Rath
License: MIT

Dataset Sources [optional]

Repository:
- kaggle dataset
- Original location (full dataset)

Uses

Direct Use

Diabetic retinopathy classification (binary or multiclass). Feature extraction (unsupervised or self supervised learning).

Out-of-Scope Use

[More Information Needed]

Dataset Structure

There are no predefined partitions in this dataset; it is up to the user to decide how to split the data.

Dataset Creation

Curation Rationale

Resizing: The images were resized to 224x224.

Source Data

Data Collection and Processing

From the description of the dataset we know that Aravind technicians travelled to rural areas in India to capture the images.

Who are the source data producers?

Aravind Eye Hospital.

Annotation process

A clinician has rated each image for the severity of diabetic retinopathy on a scale of 0 to 4:

0 - No DR

1 - Mild

2 - Moderate

3 - Severe

4 - Proliferative DR

Personal and Sensitive Information

[More Information Needed]

Bias, Risks, and Limitations

This dataset only contains a subset of the original dataset, the training split. The images have been resized by Sovit Ranjan Rath.

Recommendations

Users should be made aware of the risks, biases and limitations of the dataset. More information needed for further recommendations.

Citation

Karthik, Maggie, and Sohier Dane. APTOS 2019 Blindness Detection. https://kaggle.com/competitions/aptos2019-blindness-detection, 2019. Kaggle.

Glossary

[More Information Needed]

More Information

[More Information Needed]

Dataset Card Authors

bumbledeep