ACRIMA: Glaucoma Optic Disc Fundus Database
705 optic disc-centred fundus photographs (396 normal + 309 glaucoma) from Hospital Clinico San Carlos. Expert-annotated binary glaucoma classification.
705 optic disc-centred fundus photographs (396 normal + 309 glaucoma) from Hospital Clinico San Carlos. Expert-annotated binary glaucoma classification.
De-identified anterior-uveitis corneal-endothelial parameters and related variables.
Peripapillary RNFL and segmented-macular OCT measurements in optic-neuritis cohorts and healthy controls.
1,168 anterior-segment OCT images of keratitis with per-pixel lesion, cornea, and iris segmentation labels. Enables 3D reconstruction from B-scan stacks.
Human eye-tracking observations from an ASD study.
Paired left- and right-eye fundus images.
Human CASIA2 anterior-segment OCT repeatability measurements.
Human anterior-segment OCT keratometric-astigmatism measurements.
Human cataract-surgery biometry and clinical outcome measurements.
Bilateral LenStar biometrics from a human cataract population.
Central retinal vessel-trunk and OCT-related measurements in open-angle glaucoma eyes.
1,345 fundus images captured across multiple camera manufacturers with OD/OC segmentation + glaucoma classification labels.
Human fundus and carotid intima-media-thickness measurements.
Swept-source OCT volumes for AMD and DME with three-dimensional pigment epithelial detachment and intraretinal-fluid masks.
100 infant eyes with 491 image pairs annotated with ground-truth control points and vessel masks. Enables pediatric / ROP-relevant fundus registration research.
Human corneal-nerve measurements in Parkinson disease study participants.
Human POAG genotype observations.
Human ocular-biometry measurements from dual Scheimpflug and swept-source OCT instruments.
Wide-field fundus images from premature infants for ROP staging (Stages 1–5 + Plus Disease). Expert-annotated for AI-assisted ROP diagnosis and screening.
Clinically validated color fundus photographs from routine ophthalmic examinations with expert disease annotations.
A set of 2,246 fundus images with continuous mean opinion scores from 0 to 100, three quality grades, and the individual scores of six ophthalmologists.
800 high-resolution (2048x2048) fundus images with pixel-wise vessel segmentation. Covers normal, DR, AMD, glaucoma.
40 patients with paired pre-operative fundus images and intra-operative biomicroscopy video clips. Annotated for optic disc and vessel segmentation across domains.
1,099 pediatric fundus images from 483 premature infants annotated for retinopathy of prematurity (ROP) staging. Standard fundus (RetCam) — not ultra-widefield; `rop_uwf_intelligent` is retained as th
100 high-resolution color fundus images with pixel-wise artery / vein / crossing labels from a mixed disease cohort.
De-identified per-patient glaucoma fundus/OCT images with processed labels and annotated spreadsheet.
263 eyes × 1,115 visits. Multi-modal: VF (HFA 24-2), fundus photographs, OCT RNFL, IOP, CCT. Labels: VF progression, OD segmentation, glaucoma stage.
Time-series bright-field imaging of human embryonic-stem-cell-derived retinal organoid aggregates.
Human embryonic-stem-cell-derived RPE electrophysiology data.
4,011 color fundus images labeled for high-myopia vs pathological-myopia classification — largest public dataset for HM/PM separation.
9,939 color fundus images from Japanese patients with DR grading labels (Davis grading scale).
198 fundus images (from MESSIDOR) with pixel-wise segmentation for 10 biomarkers (lesions + structures) + DR and ME severity grades.
6,272 AS-OCT images + 392 anterior-segment photographs for corneal opacity assessment with expert grading. First large public multimodal AS-OCT + photo dataset for cornea.
Multi-modal DR dataset combining color fundus photographs (CFP), OCT B-scans, and ultra-widefield (UWF) fundus images, annotated for DR detection and severity grading. Published in Nature Scientific D
Human clinical ocular-manifestation data from four South-Kivu health zones.
Multi-source heterogeneous retinal fundus image-quality assessment dataset for training and evaluating quality-control models.
1,693 color and 1,739 infrared ocular-surface images from five clinical centers with pixel-level tear-meniscus segmentation. First multi-center dry-eye imaging benchmark.
Tobii Pro Glasses 3 natural-scene viewing recordings with gaze, IMU, event, video, and snapshot files.
1,672 OCT B-scans from AMD, DME, and healthy controls with 5,016 multi-grader layer annotations + lesion detection labels.
Anonymized South Korean cohort clinical, serologic, electrophysiologic, thymic, and treatment data.
16,530 ultra-widefield fundus images from 8,405+ patients (age 0–90) annotated for 38 ophthalmic diseases and 67 fundus features. Released alongside the FairerOPTH study on sexism and ageism in ophtha
3,859 OCT B-scan images (125 eyes, 119 patients) for macular hole segmentation. Each PNG is side-by-side: left half raw B-scan, right half colour-coded mask for 4 classes: Retina, Macular Hole, Intrar
700 Optos ultra-wide-field fundus images with multi-disease classification labels and image-quality flags from real-world clinical practice.
Ophthalmic visual question-answering benchmark dataset released as supplementary data for evaluating multimodal language models in ophthalmology.
Ophthalmology-oriented WeChat article metadata and image-link dataset for visual question answering and multimodal language-model evaluation.
Processed outer-retinal thickness and reflectance measurements from a Parkinson disease directional-OCT study.
Retinal structure, electrophysiology, and visual-perception measurements from Parkinson disease and comparison participants.
Paired color-fundus/OCT images with expert MNV subtype labels and treatment-response follow-up.
Human eye-tracking observations from a visual-preference study.
De-identified human glaucoma-surgery clinical records with postoperative ptosis and strabismus outcomes.
Human synchronized pupillometry, EEG, and fMRI salience measurements.
Pediatric ultra-widefield fundus images with professional artery-vein segmentation annotations.
Retinal vessel analysis benchmark with vessel masks, artery/vein masks and skeletons, bifurcation points, vascular trees, and abnormality annotations.
Human retinal optic-flow and locomotion observations.
Human retinal photoplethysmography and retinal-nerve-fiber-layer measurements.
Color fundus photographs paired with biological metadata and structured descriptions for retinopathy of prematurity.
3,012 OCT B-scans from 146 eyes / 130 patients with retinal vein occlusion. Dual-task labels: fluid segmentation + retinal-layer segmentation. Detection of macular lesions.
Slit-lamp photographs with expert epiphora-severity labels.
Anonymized slit-lamp photographs with English clinical annotations and vision-language conversations.
Longitudinal OCT angiography projection maps with vessel labels, clinical text, and treatment/follow-up groupings.
Human pupil and light-exposure measurements.
Human corneal-ulcer clinical images and labels.
1,219 color fundus images with DR severity grading + pixel-level exudate segmentation masks. Multi-center (SUSTech + Sun Yat-sen University).
T2-weighted orbital MRI volumes with multi-organ segmentation labels (extraocular muscles, optic nerve, lacrimal gland, orbital fat) and paired clinical data. First public orbital MRI benchmark for th
1,630 Optos ultra-wide-field fundus images from 809 patients graded for diabetic retinopathy (5-class ICDR) by senior ophthalmologists.
2,031 ultra-widefield fundus images for AI-assisted intraocular tumor detection and classification. Six categories: Normal, Choroidal Hemangioma (CH), Retinal Capillary Hemangioma (RCH), Choroidal Ost
Human visual-field perception observations.