Pixel & Pupil
An evidence series on AI and new technology in eye care.
Pixel & Pupil is evidence commentary on AI and new technology in optometry and ophthalmology, written by a practising optometrist. Each issue reads a set of published studies for what they measured, with one thread running through the series: accuracy in a validation study is not a better outcome in primary care.
A new issue every four to six weeks.
Issues
- Issue 4
Where AI now sits in an eye examination
The AI tools now reaching optometry and ophthalmology, what each one does at each point of the eye examination, and which are on the Australian register.
- Issue 1
AI is already changing medicine. What has been shown to help?
AI renewing prescriptions in Utah, 1,614 devices on the US list, general models that pass medical exams, and the trials that measured what happened to patients.
- Issue 3
Diabetic retinopathy AI: what held outside the trial, and who came back
How the autonomous diabetic retinopathy systems work in a diabetes clinic, and what studies outside their pivotal trials found at each step of the visit.
- Issue 2
The model that was right in the laboratory and wrong in the waiting room
A glaucoma model that was accurate in validation and much less so in general practice, and how little of the trial evidence on AI in eye care measures vision.
Pixel & Pupil is written by Ankit Mathur and published by OptoGuide Pty Ltd. It is not advertising, carries no testimonials and accepts no sponsorship. Figures are drawn from the published values cited beneath each one.
About OptoGuide
OptoGuide is rule-based clinical decision support for therapeutically endorsed optometrists. It is live in Australia and was built by the author of this series.
It runs alongside a consultation, from recognising a presentation through prescribing and referral, and every rule in it is deterministic.