I would say the beauty of the iris is one of the main reasons I became so interested in ophthalmology.
Every patient has their own color, texture, and tiny details in the iris—and I honestly never get tired of looking at them through the slit lamp!
That means privacy needs to be considered from the beginning:
data sharing, access control, de-identification, storage, model development, and even how images and derived features are reported.
Opening the science should not mean ignoring biometric risk.
What interests me most is that the iris may not behave like a rapidly changing biomarker.
I’m more curious whether its persistent characteristics encode aspects of genetics, pigmentation biology, development, ancestry-related variation, or perhaps biological aging.
And that is exactly the interesting question:
What biological information is hidden in the iris that we have not systematically measured yet?
Hopefully this infrastructure can eventually help explore connections between the eye and non-ocular biology too.
I think about an AI-native operating system for the ophthalmology ward.
If AI is going to change healthcare, we should first think about reality itself—and how we can encode that reality into the digital world.
The raw real-world data should also remain available—not only the extracted findings.
The central intelligence can then see the patient as one complete story rather than a collection of disconnected examinations.