Skip to main content

ROSE: Retinal OCT-Angiography Vessel Segmentation

229 OCTA images (ROSE-1 + ROSE-2) with pixel-level retinal vessel segmentation ground truth.

At a glance

FieldValue
Short namerose
Full nameROSE: Retinal OCT-Angiography Vessel Segmentation
Primary categoryocta
Contained modalitiesocta
Taskssegmentation
Samples229
Classes2 (background, vessel)
Splitstrain, test
Size0.8 GB
Source-stated termsCC BY 4.0
Normalized termscc-by
Descriptive screening labelStandard label without an explicit NC clause; not a permission finding
Terms scopedataset_files
Access frictioncontrolled_or_manual
Route backendZenodo
Availabilityavailable (checked 2026-07-21)
Acquisition supportmanual_access_blocked
Legacy sample-loader statusMetadata and access only

Notes

Zenodo mirror (2024) publishes ROSE-1 + ROSE-2 under CC BY 4.0 — upgraded from the previous form-gated academic-only distribution on imed.nimte.ac.cn. EyeDataHub auto-downloads from Zenodo record 12775880.

Access preflight and acquisition

# This route requires upstream human action; no transfer starts.
eyehub download rose --data-dir ./data --dry-run --json
# Follow the official instructions shown by preflight.

Upstream page: zenodo.org/doi

Source-term evidence: zenodo.org/doi

Loader status

This catalog record provides metadata and access instructions, but it does not yet include a standard DatasetSample loader. Inspect the source file structure or contribute a loader before using it in a training pipeline.

Citation

@misc{rose,
title = { ROSE: Retinal OCT-Angiography Vessel Segmentation },
note = { Ma et al., 'ROSE: A Retinal OCT-Angiography Vessel Segmentation Dataset and New Model', IEEE TMI 2021. Zenodo mirror: doi:10.5281/zenodo.12775880 (CC BY 4.0) },
year = { 2021 },
url = { https://zenodo.org/doi/10.5281/zenodo.12775880 },
}

Source-stated terms

  • Raw source string: CC BY 4.0
  • Normalized category: cc-by
  • Apparent scope: dataset_files
  • Descriptive screening label: Standard label without an explicit NC clause; not a permission finding

⚠️ Source-stated terms, scope, and normalized labels are curation metadata, not legal advice or a permission finding. Review the current official source before transfer or reuse.

  • octa_macula_coronal: OCTA Macula Coronal Views (82,560 records, cc-by)
  • octa_500: OCTA-500: Large-scale OCTA Multi-task Benchmark (500 records, research-only)
  • aroma_octa: AROMA Retinal OCTA Artifact Dataset (281 records, cc-by)
  • soul_octa: SOUL: OCTA Human-Machine Collaborative Annotation Dataset (178 records, cc-by)
  • drac22: DRAC 2022: Diabetic Retinopathy Analysis Challenge (174 records, cc-by)
  • ut_fsocta: UTHealth Fundus and Synthetic OCTA Dataset (count not reported records, unknown)