AIDK: anterior segment optical coherence tomography (AS-OCT) of keratitis with cornea, lesion and iris masks

Provider
Yiming Sun, Peifang Xu et al. (Zhejiang University); mirrored by MedOtter
License
CC0-1.0 (source)
Size
5 files
Format
Parquet (in sampled files)
Stored at
Hugging Face (https://huggingface.co/datasets/MedOtter/AIDK)
Access
Free and public. No account with the data provider is needed.
DOI
10.6084/m9.figshare.25952845.v1
Category
Health & Biomedical
Source
https://doi.org/10.6084/m9.figshare.25952845.v1

About this dataset

AIDK is an open set of 1,168 anterior segment optical coherence tomography (AS-OCT) images, cross-sectional scans of the front of the eye, from 64 patients with infectious keratitis (corneal infection) at the Second Affiliated Hospital of Zhejiang University. Each image has a mask with three structures: cornea, keratitis lesion and iris (labels 1 to 3, 0 background), made to support lesion segmentation and 3D reconstruction of the cornea. Images and masks are Parquet under data/, about 1.3 GB (measured 2026-11-21); read with Hugging Face datasets or pandas plus Pillow. Use for corneal lesion segmentation, keratitis severity measurement and anterior segment anatomy models. Mirrored by MedOtter from the authors' figshare deposit released under CC0 1.0, DOI 10.6084/m9.figshare.25952845.v1; cite Sun et al. 2024 and the companion Scientific Data descriptor.

How to load this dataset in Python

from datasets import load_dataset
ds = load_dataset("MedOtter/AIDK")
print(ds)

Data structure

File types in sampled files: Parquet

Files

Example file: data/train-00000-of-00003.parquet

Analyze this dataset without downloading it

On Scigantic, AIDK: anterior segment optical coherence tomography (AS-OCT) of keratitis with cornea, lesion and iris masks opens in a hosted JupyterLab notebook with the data already mounted, so you can read files where they are stored instead of copying them first. Each notebook includes an AI assistant that can see the mounted files and write and run analysis code. Open the dataset page to launch a notebook, or browse the full catalog.

This page is also available as Markdown and as JSON.

Related Health & Biomedical datasets