COph100: infant retinal fundus images for retinopathy of prematurity (ROP) registration, with vessel masks and control points

Provider
Y. Hu, M. Gong, Z. Qiu et al.; fundus images by Timkovic et al. (University Hospital Ostrava); repackaged by MedOtter
License
CC-BY-4.0 (source)
Size
3 files
Format
Parquet (in sampled files)
Stored at
Hugging Face (https://huggingface.co/datasets/MedOtter/COph100)
Access
Free and public. No account with the data provider is needed.
DOI
10.6084/m9.figshare.27061084.v1
Category
Health & Biomedical
Source
https://doi.org/10.1038/s41597-025-04426-w

About this dataset

COph100 is a fundus image registration benchmark built from 100 eyes of premature infants screened for retinopathy of prematurity (ROP) at University Hospital Ostrava. Each eye was photographed with a RetCam wide-field camera at two to nine examination sessions, so the same retina can be aligned across time as it develops. This repackaging holds 324 image sessions as 640x480 RGB fundus photographs, each paired with a binary retinal blood vessel mask (produced automatically by the authors' SS-MAF segmentation model, so treat it as soft ground truth) and the original LabelMe JSON with 10 manually placed control-point pairs per image for registration evaluation. Columns give subject, eye and session indices. One Parquet file under data/, about 110 MB (measured 2026-11-21); read with Hugging Face datasets. Use for longitudinal retinal image registration, infant retinal vessel segmentation and ROP progression studies. Labels are from the authors' figshare record (DOI 10.6084/m9.figshare.27061084, CC BY 4.0); the fundus images come from the Retinal Image Dataset of Infants and ROP (figshare, CC0). Cite Hu et al., Scientific Data 12:99 (2025), doi 10.1038/s41597-025-04426-w, and Timkovic et al., Scientific Data 11:814 (2024).

How to load this dataset in Python

from datasets import load_dataset
ds = load_dataset("MedOtter/COph100")
print(ds)

Data structure

File types in sampled files: Parquet

Files

Example file: data/test-00000-of-00001.parquet

Analyze this dataset without downloading it

On Scigantic, COph100: infant retinal fundus images for retinopathy of prematurity (ROP) registration, with vessel masks and control points 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