Eyecare-100K: Multimodal Ophthalmology VQA Corpus

~102K VQA pairs derived from 58,485 images across 8 ophthalmic modalities (fluorescein angiography, ICGA, OCT, CFP, ultrasound biomicroscopy, slit-lamp, fundus auto-fluorescence, CT) covering 100+ diseases.

At a glance

Field	Value
Short name	eyecare_100k
Full name	Eyecare-100K: Multimodal Ophthalmology VQA Corpus
Primary category	multimodal
Contained modalities	fundus, fundus_angiography, fundus_autofluorescence, oct, ocular_ultrasound, external_eye, ct, text
Tasks	classification
Samples	102,000
Classes	Not reported (Not reported)
Splits	train, val, test
Size	30.0 GB
Source-stated terms	Mixed (inherits source-dataset licenses) — needs per-row check
Normalized terms	unknown
Descriptive screening label	Unknown or unclear; do not assume permission
Terms scope	mixed_components
Access friction	anonymous_direct
Route backend	HuggingFace Hub
Availability	available (checked 2026-07-21)
Acquisition support	standard_platform_supported
Legacy sample-loader status	Metadata and access only

Notes

Status: PENDING RELEASE. Largest open ophthalmology VQA corpus when published. EyecareGPT models are out (LLSuzy/* on HF) but the VQA dataset itself is not yet public as of 2026-06.

Access preflight and acquisition

CLI

# Read-only preflight
eyehub download eyecare_100k --data-dir ./data --dry-run --json

# Explicit transfer, only when preflight reports supported behavior
eyehub download eyecare_100k --data-dir ./data

Python

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

ds = REGISTRY.get_dataset('eyecare_100k')
print(preflight_dataset(ds, './data'))  # no transfer

Upstream page: github.com/DCDmllm

Source-term evidence: github.com/DCDmllm

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{eyecare_100k,
  title  = { Eyecare-100K: Multimodal Ophthalmology VQA Corpus },
  note   = { EyecareGPT: arXiv 2504.13650; ACM MM 2025. github.com/DCDmllm/EyecareGPT },
  year   = { 2025 },
  url    = { https://github.com/DCDmllm/EyecareGPT },
}

Plain text

EyecareGPT: arXiv 2504.13650; ACM MM 2025. github.com/DCDmllm/EyecareGPT

Source-stated terms

• Raw source string: Mixed (inherits source-dataset licenses) — needs per-row check
• Normalized category: unknown
• Apparent scope: mixed_components
• 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

• ophthalvqa: OphthalVQA Dataset (count not reported records, cc-by)
• x_pcr: X-PCR Ophthalmology Progressive Clinical Reasoning Benchmark (18,700 records, unknown)
• lmod_plus: LMOD+ Multimodal Ophthalmology Benchmark (32,633 records, unknown)
• mm_retinal_reason: MM-Retinal-Reason: Ophthalmology Multimodal Reasoning Dataset (count not reported records, unknown)
• angioreport: AngioReport Fundus Angiography Report Dataset (55,361 records, unknown)
• ffa_ir: FFA-IR Medical Report Dataset (47,247 records, unknown)
• jrc_multimodal_vessels: JRC Multi-Modal Retinal Vessel Segmentation (120 records, unknown)
• multieye: MultiEYE: OCT-Enhanced Fundus Multi-Disease Benchmark (58,036 records, mit)