Insight Health Data Research Hub

03/08/2026

An AI-enabled eye screening initiative is preventing blindness in isolated parts of southern Bangladesh, where most people have never been seen by an optometrist or ophthalmologist.

Through the outreach programme, trained field staff take mobile eye clinics to remote island populations, equipped with portable colour fundus imaging devices and an AI system for identifying sight-threatening eye conditions such as diabetic retinopathy. It was built using RETFound, the open-source AI foundation model for retinal and systemic disease detection developed in the lab of Pearse Keane at UCL Institute of Ophthalmology, Moorfields Eye Hospital, supported by INSIGHT.

The pilot project is a community-led partnership of clinicians, data scientists and NGOs, conceived by Hasib Rahman of Data Yakka and Randall Stafford of Stanford University. Using RETFound, the team in Bangladesh fine-tuned an AI tool using local data, with guidance from Pearse's lab.

Read the full story 🔗

An artificial intelligence (AI) eye screening initiative, built on open-source technology developed at University College London and Moorfields Eye Hospital NHS Foundation Trust, is bringing essential eye care to isolated island communities in southern Bangladesh

25/06/2026

Diseases affecting the front of the eye (anterior segment), including cataract, keratoconus, and loss of corneal transparency, are leading causes of blindness globally, yet fewer than 10% of ophthalmic imaging datasets include them.

In response, researchers at Moorfields Eye Hospital and UCL Institute of Ophthalmology have developed CADMUS, the world's most comprehensive anterior segment dataset.

CADMUS contains nearly 1 million images linked to clinical records from 22,482 patients attending routine appointments at Moorfields. The anonymised, longitudinal dataset was built through the INSIGHT hub at Moorfields, and is being made accessible to the research community to accelerate research.

The work was led by Dr Shafi Balal and Professor Bruce Allan, supported by National Institute for Health and Care Research, NIHR Biomedical Research Centre: Moorfields, The European Society of Cataract and Refractive Surgeons (ESCRS), and T.F.C. Frost Charitable Trust.

CADMUS was developed in collaboration with Zaid Alsafi, Marcello Leucci, Pearse Keane Daniel Gore, Nikolas Pontikos

Thanks to support from NIHR and Moorfields Eye Charity CADMUS is already being used to train the first AI foundation model for the anterior segment, with potential for identifying systemic conditions.

Read more about CADMUS and researcher access: 🔗 https://www.insight.hdrhub.org/post/landmark-dataset-to-accelerate-research-on-anterior-segment-eye-disease

Friends of Moorfields

11/06/2026

BBC News Arabic journalist Somaya Bryant recently interviewed INSIGHT director Pearse Keane for a feature on the direction of artificial medical intelligence. The reporting includes perspectives on AI development and adoption in the Middle East. An English translation has kindly been provided by the author.

Read the feature in English:
https://www.insight.hdrhub.org/post/bbc-arabic-on-how-how-ai-is-revolutionising-healthcare

View the original article on BBC News عربي https://www.bbc.com/arabic/articles/cwy2nvq111go

Moorfields Eye Hospital, UCL Institute of Ophthalmology

بدأت تقنيات الذكاء الاصطناعي بالفعل في إحداث تحولات كبيرة في مجال الرعاية الصحية في مختلف أنحاء العالم، بما في ذلك المنطقة العربية. لكن الخبراء ينبهون إلى ضرورة وجو...

02/06/2026

Thank you to the incredible Friends of Moorfields volunteers who make such a positive difference to the experience of patients visiting Moorfields Eye Hospital 👏

20/05/2026

Our colleagues in outback Western Australia are using AI as a clinical tool to stop people going blind from diabetic retinopathy (DR). This is a region where residents may have to travel up to 3,000 km to reach a medical centre. Instead, a mobile AI screening service goes to people in remote areas, helping them to get diagnosed and treated sooner.

Led by Angus Turner with Mark Chia at Australian non-profit organisation Lions Outback Vision, the breakthrough medical AI service was developed in collaboration with First Nations Australians, who are disproportionately affected by DR.

The project was delivered with technology partners Google and Topcon Healthcare, supported by Pearse Keane's lab at UCL Institute of Ophthalmology and the INSIGHT Hub at Moorfields Eye Hospital.

Read how the work evolved 🔗https://www.insight.hdrhub.org/post/novel-ai-system-tackles-eye-health-inequalities-in-outback-australia

Watch the video to find out about the project's impact ⬇️ https://youtu.be/h_W9dZjdii0?si=-GuCqdPGeSTfaNFb

Join in Broome, Western Australia, as he meets with eye...

19/05/2026

What can we expect from the next wave of medical AI? Writing for the College of Optometrists, journalist Helen Bird explores how and multiomics could shape the future of eye care.

The feature investigates how linkages between eye data and other biological data such as genomics are advancing the role of ocular biomarkers in unlocking complex biological systems - informed by interviews with oculomics pioneer Pearse Keane and Professor Anthony Khawaja of Moorfields Eye Hospital UCL Institute of Ophthalmology, INSIGHT, Cascader.

Read the article 🔗
https://www.insight.hdrhub.org/post/oculomics-multiomics-and-the-future-of-eye-health

Reproduced with kind permission of the editors of Acuity.

11/05/2026

Next Wednesday 20th May, the National AI Commission is holding an online session to answer questions from the public on the regulation of AI in healthcare ⬇️

We’d like to set additional cookies to understand how you use GOV.UK, remember your settings and improve government services.

21/04/2026

Researchers at Tohoku University and UCL have developed a more precise AI model for measuring retinal age gap - a biomarker indicating the difference between a person’s biological age and their chronological age.

In addition to training and validating the model on large-scale eye imaging datasets from Japan and the UK, the team added the blood sugar marker HbA1c to the training data, improving the model’s accuracy.

The model was able to predict age from a single fundus image, with accuracy of around 3 years on average, which is an improvement on previous benchmarks. It showed that a larger gap is associated with diabetes, heart disease, and stroke.

Because the model works from a single routine eye photograph, with no blood tests required at the point of use, it could be developed for clinical settings to flag people with a large retinal gap, so they can be referred for systemic health checks.

The paper is published in Communications Medicine, and is part of a wider collaboration between Tohoku and UCL.
bit.ly/4mJx7Hm

A picture (of the eye) is worth a thousand words to this AI system that predicts the risk of developing major diseases like diabetes.

14/04/2026

A portable GPU could power research institutions in Rwanda and other Low Middle Income countries (LMICs) with the computing power they need to contribute training datasets to the world’s first globally representative medical AI foundation model, Global RETFound.

The initiative is co-led by UCL Institute of Ophthalmology, Moorfields Eye Hospital, INSIGHT Eye Hub, National University of Singapore Yong Loo Lin School of Medicine, and The Chinese University of Hong Kong 香港中文大學 - CUHK.

Bringing together more than 100 research groups to contribute data to the model is a technical challenge. Not all groups have the infrastructure for easy curation of local, anonymised patient datasets.

To find a way forward, our colleague Fiston Gatera recently travelled to Rwanda with a portable GPU, successfully testing it with a team at CIIC-HIN (SEEK-IN), a centre of the Africa Knowledge Institute for Innovation and Scientific Advancement (AKIISA).

The plan is to enable centres in LMICs with no GPU to collaborate on building the Global RETFound foundation model. This will enhance equity as well as capacity building in medical AI in LMICs.

Read more about Global RETFound: https://www.insight.hdrhub.org/post/quest-to-build-the-first-medical-ai-foundation-model-with-globally-representative-data

08/04/2026

We have updated our data catalogue for (DR), among the world's most comprehensive and diverse longitudinal datasets curated for this condition.

DR affects a third of the 4.4 million people estimated to have diabetes in the UK, and is a leading cause of preventable blindness in the adult working population globally.

Through INSIGHT at Moorfields Eye Hospital, DR data is routinely collected from patient appointments, anonymised and curated into information-rich datasets. This data is accessible to researchers working for patient benefit, and is a valuable bio-resource representing the diversity of the population served by Moorfields.

INSIGHT data and advanced infrastructure supports much of the research work undertaken in the medical lab of Pearse Keane across Moorfields and UCL Institute of Ophthalmology, in addition to external projects in the wider research ecosystem.

For more information view the recently updated INSIGHT DR data catalogue
🔗 https://lnkd.in/eyf-UHxC

Find out how to apply for access
🔗 https://lnkd.in/eT7QZx8k