AI in ophthalmology is expanding fast
Hundreds of new AI/ML models are published every year across diabetic retinopathy, glaucoma, cataract, oculoplastics, and beyond. Without a structured map, the literature becomes harder to navigate, not easier.
EyeModelHub is a living scoping review and model-card registry for AI/ML in ophthalmology — peer-screened, APPRAISE-AI graded, dated.
The literature on AI in ophthalmology grows faster than any single clinician or researcher can read it. EyeModelHub systematically maps every published AI/ML model to its imaging modality, model family, validation cohort, regulatory status, and methodological quality.
Built around a living-review framework expanded to nine subspecialties — annual full re-searches per domain, continuous user submissions, and every included study graded with the APPRAISE-AI rubric.
Four reasons EyeModelHub exists.
Hundreds of new AI/ML models are published every year across diabetic retinopathy, glaucoma, cataract, oculoplastics, and beyond. Without a structured map, the literature becomes harder to navigate, not easier.
Most published ophthalmic AI is single-center, never independently validated, and rarely tested against demographic subgroups. EyeModelHub surfaces these gaps as first-class data on every card.
CONSORT-AI, TRIPOD+AI, CLAIM, DECIDE-AI — the field has mature reporting frameworks but inconsistent uptake. APPRAISE-AI scoring exposes which papers actually meet them.
When an oculoplastic surgeon, retinal specialist, or AI researcher asks "what is the state of the art for this task?", they should be able to answer in seconds — not hours.
A four-tier editorial structure. Director, Associate Directors, and nine Subspecialty Leads pictured; reviewers acknowledged below.
Conceived and founded EyeModelHub. Leads editorial direction, schema design, web architecture, methodology, and stakeholder engagement across all nine subspecialties.
Owns the protocol, schema versioning, controlled vocabularies, and the APPRAISE-AI rubric.
Stakeholder engagement, institutional partnerships, conference presence, and external communication.
Maintains the reviewer roster, runs APPRAISE-AI calibration, and audits inter-rater agreement across subspecialties.
Diseases of the eyelid, orbit, lacrimal system, and periocular tissues — reconstructive, oncologic, and aesthetic.
Disorders of the neurosensory retina, vitreous, and choroid — vascular, degenerative, inflammatory, and inherited.
Progressive optic neuropathy driven by intraocular pressure and structural change — irreversible without intervention.
External diseases of the cornea, conjunctiva, and ocular surface — including transplantation and ectasia.
Visual manifestations of central nervous system disease — afferent and efferent visual pathway disorders.
Childhood eye disease and ocular motility — amblyopia, strabismus, and disorders of visual development.
Benign and malignant tumours of the eye, orbit, and adnexa — both intraocular and periocular disease.
Lens opacification and refractive correction — cataract surgery, IOL selection, and laser refractive procedures.
Histopathology of ocular and orbital tissue — tumour, inflammatory, and degenerative disease.
Co-authors on one or more subspecialty reviews who contributed to title/abstract screening, full-text review, data extraction, or APPRAISE-AI scoring.
For collaboration, methodology questions, partnerships, press, or model suggestions. Routed directly to the editorial team.