LMOD+ Multimodal Ophthalmology Benchmark

Composite multimodal ophthalmology benchmark with multi-granular anatomical, diagnostic, staging, demographic, and text annotations.

At a glance

Field	Value
Short name	lmod_plus
Full name	LMOD+ Multimodal Ophthalmology Benchmark
Publication date	Unknown
Date basis	Unknown
Publication date precision	Unknown
Publication date evidence	Unknown
Publication date source field	Unknown
Publication date reviewed	Unknown
Date notes	-
Primary category	multimodal
Resource role	derivative_dataset
Dataset family	lmod_plus
Contained modalities	fundus, oct, external_eye, surgical_video, text, tabular
Tasks	visual_question_answering, classification, grading, detection
Primary reported quantity	32,633 annotated instances
Classes	Not reported (Not reported)
Splits	all
Size	Not reported
Source-stated terms	Mixed upstream licenses; verify each component
Normalized terms	unknown
Descriptive screening label	Unknown or unclear; do not assume permission
Terms scope	mixed_components
Access friction	anonymous_direct
Route backend	Manual (upstream-gated)
Availability	available (checked 2026-07-21)
Acquisition support	guided_instructions_only
Legacy sample-loader status	Metadata and access only

Reported quantities

Role	Count	Unit	Scope	Basis	Evidence
Primary	32,633	annotated_instances	Composite benchmark instances	official_source_description	kfzyqin.github.io/lmod_plus

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

Integrates 32,633 instances from nine public datasets across five modalities and 12 conditions. It adds a distinct benchmark and annotation surface but does not replace the source datasets; license and redistribution terms remain component-specific.

Documented relationships

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

• This record is derived from cataract1k: The LMOD+ project page lists nine component datasets, including the five cataloged targets represented by these edges. (evidence)
• This record is derived from g1020: The LMOD+ project page lists nine component datasets, including the five cataloged targets represented by these edges. (evidence)
• This record is derived from idrid: The LMOD+ project page lists nine component datasets, including the five cataloged targets represented by these edges. (evidence)
• This record is derived from oimhs: The LMOD+ project page lists nine component datasets, including the five cataloged targets represented by these edges. (evidence)
• This record is derived from refuge2: The LMOD+ project page lists nine component datasets, including the five cataloged targets represented by these edges. (evidence)

Access information and download

CLI

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

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

Python

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

ds = REGISTRY.get_dataset('lmod_plus')
print(preflight_dataset(ds, './data'))  # no download

Upstream page: kfzyqin.github.io/lmod_plus

Source-term evidence: kfzyqin.github.io/lmod_plus

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{lmod_plus,
  title  = { LMOD+ Multimodal Ophthalmology Benchmark },
  note   = { Qin Z, Liu Y, Yin Y, et al. LMOD+: A Comprehensive Multimodal Dataset and Benchmark for Developing and Evaluating Multimodal Large Language Models in Ophthalmology. ACM Transactions on Computing for Healthcare. 2026. doi:10.1145/3801746 },
  year   = { 2026 },
  url    = { https://kfzyqin.github.io/lmod_plus/ },
}

Plain text

Qin Z, Liu Y, Yin Y, et al. LMOD+: A Comprehensive Multimodal Dataset and Benchmark for Developing and Evaluating Multimodal Large Language Models in Ophthalmology. ACM Transactions on Computing for Healthcare. 2026. doi:10.1145/3801746

Source-stated terms

• Raw source string: Mixed upstream licenses; verify each component
• 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.

Similar resources by shared modality

• eyecare_100k: Eyecare-100K: Multimodal Ophthalmology VQA Corpus (102,000 question answer pairs, unknown)
• x_pcr: X-PCR Ophthalmology Progressive Clinical Reasoning Benchmark (18,735 rows, unknown)
• ophthalvqa: OphthalVQA Dataset (600 question answer pairs, cc-by)
• fairvlmed: FairVLMed: Fair Vision-Language Medical Ophthalmic Dataset (10,000 images, cc-by-nc-nd)
• olives: OLIVES: Ophthalmic Labels for Investigating Visual Eye Semantics (9,408 b scans, cc-by)
• grape: GRAPE: Glaucoma Real-world Appraisal Progression Ensemble (1,115 examinations, cc0)
• mm_retinal_reason: MM-Retinal-Reason: Ophthalmology Multimodal Reasoning Dataset (130 question answer pairs, unknown)
• ocular_chat_vqa: OcularChat-VQA: AREDS-Derived Patient-Physician Dialogue Dataset (844,000 records, cc-by-nc-sa)