Skip to main content

Diabetic Macular Edema Visual Question Answering Dataset

Fundus-image VQA dataset for diabetic macular edema derived from IDRiD and e-ophtha.

At a glance

FieldValue
Short namedme_vqa
Full nameDiabetic Macular Edema Visual Question Answering Dataset
First published2022-06-27
Publication date precisionday
Publication date evidencezenodo.org/api
Publication date source fieldmetadata.publication_date (earliest repository version)
Publication date reviewed2026-09-11
Primary categorymultimodal
Resource roleannotation_layer
Dataset familydme_vqa
Contained modalitiesfundus, text
Tasksvisual_question_answering, classification
Primary reported quantity13,470 question answer pairs
ClassesNot reported (Not reported)
Splitsall
Size0.1 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 frictionself_service_authenticated
Route backendZenodo
Availabilityavailable (checked 2026-07-21)
Acquisition supportstandard_platform_supported
Legacy sample-loader statusMetadata and access only

Reported quantities

RoleCountUnitScopeBasisEvidence
Primary13,470question_answer_pairsTrain, validation, and test QA pairsderived_from_reported_componentszenodo.org/records
Additional679imagesTrain, validation, and test imagesderived_from_reported_componentszenodo.org/records

Counts retain their source-reported units. Additional rows can describe components, paired items, or derivative copies and are not automatically added to the primary quantity.

Notes

Derivative VQA layer built from existing fundus datasets.

Documented relationships

These links record source-supported lineage or overlap, not merely similar modality tags.

  • dme_vqa_logical is extension of this record: This release adds logical-relation annotations to the earlier DME VQA resource. (evidence)
  • This record is derived from e_ophtha: The DME VQA deposit identifies e-ophtha images as source material. (evidence)
  • This record is derived from idrid: The DME VQA deposit identifies IDRiD images as source material. (evidence)

Access information and download

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

# Download, only when preflight reports supported behavior
eyehub download dme_vqa --data-dir ./data

Upstream page: zenodo.org/records

Source-term evidence: zenodo.org/records

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{dme_vqa,
title = { Diabetic Macular Edema Visual Question Answering Dataset },
note = { Diabetic Macular Edema Visual Question Answering Dataset. Zenodo, 2022. doi:10.5281/zenodo.6784358 },
year = { 2022 },
url = { https://zenodo.org/records/6784358 },
}

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.

Similar resources by shared modality

  • eyecare_100k: Eyecare-100K: Multimodal Ophthalmology VQA Corpus (102,000 question answer pairs, unknown)
  • angioreport: AngioReport Fundus Angiography Report Dataset (55,361 images, unknown)
  • ffa_ir: FFA-IR Medical Report Dataset (47,247 images, unknown)
  • lmod_plus: LMOD+ Multimodal Ophthalmology Benchmark (32,633 annotated instances, unknown)
  • x_pcr: X-PCR Ophthalmology Progressive Clinical Reasoning Benchmark (18,735 rows, unknown)
  • deepeyenet: DeepEyeNet (DEN): Fundus Report Generation Dataset (15,709 images, research-only)
  • dme_vqa_logical: DME VQA Dataset with Logical Relations (13,470 question answer pairs, cc-by)
  • fairvlmed: FairVLMed: Fair Vision-Language Medical Ophthalmic Dataset (10,000 images, cc-by-nc-nd)