EyeQ Retinal Image Quality Assessment Dataset

Quality labels for 28,792 EyePACS fundus images, graded as good, usable, or reject and divided into the original EyePACS train and test partitions.

At a glance

Field	Value
Short name	eyeq
Full name	EyeQ Retinal Image Quality Assessment Dataset
Primary category	fundus
Contained modalities	fundus
Tasks	quality, grading
Samples	28,792
Classes	3 (good, usable, reject)
Splits	train, test
Size	Not reported
Source-stated terms	Unknown for released quality labels; code is CC BY-NC-SA 4.0
Normalized terms	unknown
Descriptive screening label	Unknown or unclear; do not assume permission
Terms scope	dataset_files
Access friction	anonymous_direct
Route backend	GitHub
Availability	available (checked 2026-07-21)
Acquisition support	loader_implemented_not_live_tested
Legacy sample-loader status	Metadata and access only

Notes

EyeQ is a distinct annotation layer over EyePACS. The repository license covers the code but does not clearly license the quality labels. It does not provide the source images. Users must obtain EyePACS separately and comply with its terms.

Access preflight and acquisition

CLI

# Read-only preflight
eyehub download eyeq --data-dir ./data --dry-run --json

# Explicit transfer, only when preflight reports supported behavior
eyehub download eyeq --data-dir ./data

Python

from eyedatahub.acquisition import preflight_dataset
from eyedatahub.datasets.registry import REGISTRY

ds = REGISTRY.get_dataset('eyeq')
print(preflight_dataset(ds, './data'))  # no transfer

Upstream page: github.com/HzFu

Source-term evidence: github.com/HzFu

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

BibTeX

@misc{eyeq,
  title  = { EyeQ Retinal Image Quality Assessment Dataset },
  note   = { Fu H, Wang B, Shen J, et al. Evaluation of retinal image quality assessment networks in different color-spaces. MICCAI. 2019. doi:10.1007/978-3-030-32239-7_6 },
  year   = { 2019 },
  url    = { https://github.com/HzFu/EyeQ },
}

Plain text

Fu H, Wang B, Shen J, et al. Evaluation of retinal image quality assessment networks in different color-spaces. MICCAI. 2019. doi:10.1007/978-3-030-32239-7_6

Source-stated terms

• Raw source string: Unknown for released quality labels; code is CC BY-NC-SA 4.0
• Normalized category: unknown
• Apparent scope: dataset_files
• Descriptive screening label: Unknown or unclear; do not assume permission

⚠️ 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.

Related datasets with shared modalities

• airogs: AIROGS: AI for Robust Glaucoma Screening (113,893 records, cc-by-nc-nd)
• eyecare_100k: Eyecare-100K: Multimodal Ophthalmology VQA Corpus (102,000 records, unknown)
• justraigs: JustRAIGS: Just Referral AI Glaucoma Screening Dataset (101,442 records, cc-by-nc-nd)
• eyepacs: EyePACS — Diabetic Retinopathy Detection (Kaggle 2015) (88,702 records, research-only)
• multieye: MultiEYE: OCT-Enhanced Fundus Multi-Disease Benchmark (58,036 records, mit)
• angioreport: AngioReport Fundus Angiography Report Dataset (55,361 records, unknown)
• ffa_ir: FFA-IR Medical Report Dataset (47,247 records, unknown)
• bidr: BiDR: Diabetic Retinopathy Diagnosis Dataset (35,126 records, unknown)