Data Use Application 016
Lead applicant organisation name
Name of the legal entity that signs the contract to access the data.
Moorfields Eye Hospital NHS Foundation Trust; University College London. Lead applicant: Shafi Balal, NIHR Doctoral Fellow, Research student PhD
Project title
The title of the project/research study request that the applicant is investigating through the use of health data.
KERAFound: an artificial intelligence foundation model for the anterior segment of the eye
Lay summary
A concise and clear description of the project. This should outline the problem, objective and the expected outcomes in language that is understandable to the general public.
The purpose of this project is to develop a first-of-its-kind artificial intelligence (AI) foundation model—KERAFound—to improve the early diagnosis and management of diseases affecting the front of the eye (the anterior segment). Some of these conditions, such as kerataconus, can lead to severe visual impairment and blindness. The project will also investigate the model's potential to recognise early signs of conditions affecting other parts of the body, such as cardiovascular disease, stroke or inflammation, which can be treated with preventative intervention.
Training effective medical AI models typically requires large amounts of data that has been reviewed by experts and labelled with information about what disease may be evident from the data - a process that can take years and requires many hours of input from experts. AI foundation models can harness the power of large amounts of data without this expensive and time-consuming step of requiring experts to label all the data. Foundation models have been proven to be more effective than traditional AI models, and because they are trained on a large amount of diverse data. This diversity of data means that the foundation model can be tailored to work effectively for different patient populations.
Public benefit statement
A description in plain English of the anticipated outcomes, or impact of project on the general public.
KERAFound will provide public benefit in three significant ways. 1. It will enable earlier and more sensitive detection of diseases affecting the front of the eye, particularly keratoconus. Kerataconus affects mainly teenagers and young adults, and early intervention is crucial to prevent vision loss. The model will also be capable of detecting early signs of infectious keratitis, an infection of the cornea that can lead to vision loss and corneal scarring, requiring a corneal transplant. By facilitating earlier intervention for progressive corneal diseases, KERAFound could contribute to reducing the global burden of preventable vision loss, with corresponding improvements in quality of life and economic productivity. 2. Successful implementation would demonstrate the feasibility of AI-augmented ophthalmology care, potentially catalysing wider digital transformation in eye care delivery. 3. The model and its visualisation capabilities could serve as valuable educational tools for training ophthalmologists and optometrists, improving understanding of disease patterns and progression.
Latest approval date
The last date the data access request for this project was approved by a data custodian.
30 May 2025
Dataset(s) name
The name of the dataset(s) being accessed.
Bespoke: anterior segment OCT imaging; Placido topography imaging
Access type
Determines whether the data will be accessed with an Trusted Research Environment (TRE) or via data release.
Data provisioned to the research applicant through INSIGHT's Secure Research Environment
Data sensitivity level
The level of identifiability of the data being accessed, as defined by Understanding Patient Data.
Anonymised