Skip to main content

Landmark eye imaging dataset set to accelerate research into sight-threatening conditions

Share:

3 min read

Themes:

Subjects:

Researchers at Moorfields Eye Hospital and UCL Institute of Ophthalmology have developed the world’s largest dataset of images showing the anterior segment (front) of the eye, creating an important new resource to support research into sight-threatening eye conditions.

Known as CADMUS, the dataset contains almost one million anonymised eye images linked to clinical information from more than 22,000 patients treated at Moorfields Eye Hospital. The data has been made available to researchers through INSIGHT, the Eye and Oculomics Health Data Research Hub.

Three side-by-side images showing different ways of examining an eye. The first image shows a close-up of the eye’s surface with circular rings reflected on it, mapping its shape. The middle image shows a cross-section view of the front of the eye taken with a scanning technique. The third image is a standard photograph of an eye looking straight ahead, showing the iris and pupil.
Images taken from the CADMUS dataset showing different examinations of the front of the eye.
Image reproduced with author permission under a Creative Commons Attribution 4.0 International Licence.

Conditions affecting the front of the eye including cataracts and keratoconus are among the leading causes of vision loss worldwide. However, researchers have had access to relatively few large datasets to help them better understand these conditions and develop new ways to diagnose and treat them.

CADMUS helps address this gap by bringing together eye scans, photographs and clinical information collected during routine patient care. Because many patients were seen over several years, the dataset will also help researchers understand how eye conditions change over time and how treatments affect outcomes.

The project was led by Dr Shafi Balal, an ophthalmic surgeon at Moorfields Eye Hospital and NIHR doctoral fellow at UCL, together with Professor Bruce Allan, consultant ophthalmic surgeon at Moorfields Eye Hospital and honorary professor at UCL.

Headshot of Dr Shafi Balal wearing a navy suit jacket, white shirt, and black tie. He has short dark hair and is positioned against a plain, light-coloured indoor background.
Dr Balal said
“Early research using CADMUS data has already produced promising results. We have used the dataset to improve understanding of how keratoconus progresses and trained artificial intelligence tools using CADMUS data. One model can predict patient age and biological sex from anterior segment scans, demonstrating that routine clinical images carry rich biological signals invisible to the human eye.”

With support from the NIHR Moorfields Biomedical Research Centre (BRC) and Moorfields Eye Charity, researchers are already using CADMUS to develop new artificial intelligence (AI) tools that could help improve the detection and monitoring of eye disease in the future.

The dataset has been published in the journal Ophthalmology Science and is available to approved researchers through INSIGHT’s Data Use Application process.

Find the CADMUS datasheet in Ophthalmology Science

By supporting resources such as CADMUS, our BRC is helping researchers harness the power of data and AI to better understand eye disease and develop new diagnostic tools to improve patient care.

Further information