Research prototypeEye checks by phone

Many eye checks. Just your phone.

Iris is exploring one 30-second phone scan that looks for signs linked to diabetic retinopathy, glaucoma and cataracts, using only your camera and screen light. It's a student research project: you can try the prototype, but it isn't a medical device and can't diagnose anything.

A woman holding her phone up at eye level, looking at the screen

Trailer

See what Iris can do

Three checks

One scan, three things to look for

Each check reuses the same light and camera. None of them can diagnose anything. They're meant to tell you when it's worth seeing a doctor, and we're honest about how far along each one is.

Two boys holding balls, with dark blotches covering parts of the picture, as someone with diabetic retinopathy might see them
How diabetic eye damage (retinopathy) can look

Simulation: National Eye Institute, NIH

01

Diabetic retinopathy

Pupil light reflex

How big your pupils are at rest, and how far, how fast and how soon they shrink after a flash.

Why it matters
Diabetic retinopathy affects the retina in people with diabetes. Iris is exploring this as a research focus.
Evidence Research question
Clinic pupillometer studies found differences in people with diabetes. These findings do not establish that Iris can detect diabetic retinopathy.
Limits
A pupil response cannot diagnose or rule out diabetic retinopathy. Iris does not capture a clinical retinal image.

If flagged: An eye exam with retinal assessment

Two boys holding balls, clear in the middle but dark around the edges, as someone with glaucoma might see them
How glaucoma can narrow your side vision

Simulation: National Eye Institute, NIH

02

Glaucoma

Left vs right reaction

How strongly your pupils react when only your left eye is lit, compared with only your right eye.

Why it matters
More than half of people with glaucoma don't know they have it. It steals side vision slowly and without pain.
Evidence Known test, new device
It's the phone version of a test eye doctors use (the swinging flashlight test). Early studies with phone pupillometers are promising but small.
Limits
It only notices damage that's worse in one eye, and other optic-nerve problems can look the same.

If flagged: An eye exam (eye pressure and optic nerve)

Two boys holding balls, blurry and hazy all over, as someone with cataracts might see them
How cataracts can blur and dim your vision

Simulation: National Eye Institute, NIH

03

Cataracts

Lens cloudiness

Whether the pupil looks hazy or gray in the close-up instead of clear and dark.

Why it matters
Cataracts are very common with age and the leading cause of blindness worldwide, and surgery can fix them.
Evidence Best studied
Several phone + AI studies matched eye doctors about 90% of the time, mostly with controlled lighting. Iris reuses the same close-ups as the other two checks.
Limits
Works best for advanced cataracts; glare and lighting can look like haze, and early ones may be missed.

If flagged: An eye exam

Only a doctor can diagnose diabetic retinopathy, glaucoma or cataracts. See the studies

How it will work

About 30 seconds, any phone

Step 1

Get close

Soft light, screen brightness up. Cover one eye and bring the phone close to the other.

Step 2

Right eye, one flash

Iris locks onto your eye and records how your pupil reacts to the flash.

Step 3

Left eye, one flash

The same for your other eye, so the two can be compared.

Step 4

Three results

Each check says “looks typical” or “see a doctor”, in plain words.

Why it matters

Ideas that guide Iris

“An ounce of prevention is worth a pound of cure.”
BF
Benjamin FranklinInventor and Founding Father1735
3planned checks
Phone onlyhardware goal
Researchcurrent stage

01 / 03

Roadmap

What we're building, and when

From research to a free app, one careful step at a time. Dates are goals and may change as we learn.

Fall 2026

Now

Research and concept

Choosing what to check and learning what the science can and can't support.

  • Picked three checks that share one phone scan: diabetic retinopathy, glaucoma and cataracts
  • Reviewed the published studies behind each one
  • Built this website, the app preview and the Ask Iris assistant

Winter 2026

Next

Phone prototype

Turning the plan into a scan that runs on a real phone, with images processed on the device.

  • Track both pupils with the front camera
  • Run the full scan: one flash for both eyes, then one per eye
  • Measure lens haze in the flash photos
  • Try it on different phones, rooms and eye colors

Early 2027

Planned

Accuracy checks

Checking the phone against the tools eye doctors already trust.

  • Compare pupil readings with a clinic pupillometer
  • Compare left vs right results with the swinging flashlight test
  • Compare haze scores with an eye doctor's exam
  • Repeat scans to see how consistent they are

Spring 2027

Planned

Pilot study

A small study with volunteers, run with permission and adult supervision.

  • People with and without clinically confirmed diabetic retinopathy, matched by age
  • People with glaucoma or cataracts diagnosed by a doctor
  • Share what worked and what didn't, openly

Later 2027

Planned

Free beta app

Only if the pilot shows the checks are useful and reliable.

  • Plain-language results and when to see a doctor
  • Help finding low-cost care in Jacksonville
  • Images stay on your phone

About

A student project with a simple goal

Millions of people live with health problems they don't know about yet. Iris is a team of students exploring whether the phone already in your pocket could flag early warning signs through your eyes, and point you toward a doctor sooner.

Iris is a research prototype, not a medical device. It will never diagnose anything; it's meant to help people decide when to get checked.

Meet the team

Privacy first

Camera images are planned to stay on your phone.

Research led

Built on published, peer-reviewed eye studies.

No extra hardware

Just the phone you already have.

FAQ

Frequently asked questions

The three planned checks, how one scan covers them, and what a result could and couldn't tell you. Hover a row, or use the button below, to pause.

What are the three checks Iris is working on?

Iris is exploring three checks in one phone scan: pupil responses being explored for diabetic retinopathy research, differences between the two eyes that may relate to glaucoma, and lens cloudiness associated with cataracts. All three are in development. The goal is to help people understand when to seek professional care, and each check needs its own testing.

How will one scan cover all three?

The plan is to line up both eyes in a dim room, record their response to the screen light, then light each eye separately while the other is covered. Iris would use the pupil response for the diabetic retinopathy research, compare the separate responses for the glaucoma research, and examine captured images for cataract-related cloudiness. We're aiming for a short scan of around 30 seconds; the final steps and timing still need phone testing.

Can the pupil check tell me if I have diabetic retinopathy?

No. Iris has not established that pupil measurements can detect diabetic retinopathy. Studies of pupil responses in diabetes do not validate a retinopathy test. An eye professional needs to assess the retina; a blood sugar test alone does not diagnose diabetic retinopathy.

How will the glaucoma check work?

Iris plans to compare the pupil responses when each eye is lit separately. A weaker response from one eye can occur with optic-nerve problems, including glaucoma, but it can have other causes too. Early glaucoma, or damage affecting both eyes similarly, may be missed. This check would suggest an eye exam; it won't measure eye pressure or diagnose glaucoma.

What will Iris look for with cataracts?

A cataract is cloudiness in the eye's lens. Iris plans to examine images captured during the scan for visible haze or changes in the light reflected through the pupil. Glare and lighting can look similar, and early cataracts may not be visible in a phone photo. An eye doctor would need to confirm any finding with an eye exam.

Is eye scanning available yet?

Yes, as a research prototype. After signing in, you can try an eye photo check and an experimental pupil test. It isn't a medical device, and no result can diagnose or rule out disease. We still need to test repeatability, more phones and eye colors, and compare each check with appropriate clinical assessments. There isn't a launch date for a finished app.

Will I need extra equipment or a special phone?

The goal is to use only a phone's camera and screen light, with no lens attachment or separate device. We haven't confirmed which phones and browsers will support all three checks reliably. Dim lighting, a steady phone, and clear pupil images are part of the planned capture process.

Will my eye images be uploaded?

The plan is to process eye images on your phone without uploading them. The AI explanation may use measurement numbers rather than images. The separate Ask Iris chat sends your messages to an AI provider, and chat history is saved in your browser. The Privacy Policy explains the current website's data handling.

What would a typical or flagged result mean?

A typical result would only describe what the prototype measured; it wouldn't rule out diabetic retinopathy, glaucoma, or cataracts. A flagged or unclear result would be a reason to discuss appropriate testing with a healthcare professional. Iris is a research prototype and cannot replace blood tests, regular eye exams, or medical advice.

How could Iris help the Jacksonville community?

We're exploring whether an accessible phone check could encourage people to seek an eye exam sooner, especially when cost, transportation, or time make care harder to reach. The plan includes listening to local residents and healthcare workers and helping people find care resources. That community benefit depends on demonstrating that the checks are useful and reliable.

Still curious? Explore the research or ask the Iris assistant in the corner.