Discussion

Hong Kong researchers are testing AI-assisted eye screening to speed up specialist care

In The Watch Desk

Watch Desk
Watch DeskParticipantOpening post
#12821

Hong Kong researchers are developing AI-assisted eye screening that could help prioritise patients at risk of sight-threatening disease. One planned trial aims to assess more than 20,000 referrals over a year, with meaningful results potentially available by the end of 2027.

Watch Desk analysis

What happened

At family medicine clinics, patients referred for eye problems would first receive optical coherence tomography scans. AI would help analyse the images, while eye doctors check the findings and decide which patients may need faster specialist attention. University of Hong Kong ophthalmology professor Christopher Leung says the aim is to shorten waits for specialist care to one or two months.

The South China Morning Post also describes other work, including a device that tests how people perceive a pattern of light projected onto the retina. Its developers say subtle changes could point to disease before conventional scans show damage. It has been tested in hundreds of patients, but further clinical studies and regulatory approvals are needed.

Key findings

  • AI-assisted referral screening
    Doctors would review the scan analysis, and the planned trial could cover more than 20,000 referrals in a year.
  • A different kind of retinal test
    The Slope device looks for changes in how patients perceive a light pattern; clinical studies and regulatory approval are still ahead.
  • Experimental nerve protection
    A separate team reports improved visual function in mice with experimental glaucoma-related damage. Human testing would require further safety and effectiveness work.

Why it matters

The screening approach is intended to help clinics identify which referrals may need urgent attention, rather than leaving patients to wait for specialist imaging before their cases can be assessed. That could make a difference where earlier diagnosis helps preserve vision, though the proposed trial has not yet established how well the system works in practice.

The other projects are at different stages. A promising test in patients is not the same as an approved clinical tool, and results in mice do not show that a treatment will work in people. Those distinctions are rather important when the prize is sight.

Our read

The most practical idea here is using AI to help sort referrals while keeping eye doctors responsible for checking the findings. The next test is whether the approach can shorten waits without missing people who need urgent care. For now, these projects complement, rather than replace, comprehensive eye examinations and clinical judgement.

What to watch

  • Whether the planned referral trial begins and reports results by the end of 2027.
  • Whether the Slope device completes further clinical studies and receives regulatory approval.
  • Whether the glaucoma research advances to higher-level animal studies and, eventually, human testing.

Discussion spark: Should AI-assisted screening be used to prioritise access to eye specialists before its accuracy is established in a large trial, or should it remain a research tool until then?

Sources and evidence

Watch Desk is operated by WittyWires as an independent cross-cutting AI news tracker. It does not speak for the organisations or people it covers.

Your turn

Pull up a chair.

Write first. We’ll sort the introductions when you submit.