Pearse Keane

About Pearse Keane

What major industry/global trends are catching your attention right now?

The advances we are seeing in the capabilities of artificial intelligence (AI) foundation models are very intriguing. In parallel with the rise of vision-language models, the development of “reasoning” models, which take longer to answer but give more structured responses based on a logical, step-by-step approach, is particularly significant. In the multi-disciplinary group I lead across UCL Institute of Ophthalmology and Moorfields Eye Hospital, we have been exploring the potential of some of these reasoning models for research tasks, and for creating personalized tools for clinicians without requiring any coding knowledge. It will be interesting for us to leverage some of these new approaches to support our own work in foundation models, refining model architecture, and optimizing the structure of training data for even better performance.

In what ways do you think AI and machine learning will impact ophthalmic innovation?

One of the most significant impacts will be in democratizing access to eye care. In regions with limited specialist availability, AI-driven tools can extend high-quality screening and diagnosis to populations who may not have easy access to healthcare yet have high need. This democratization of care has the potential to significantly reduce preventable blindness worldwide.

For example, we have worked with Australian colleagues in developing an AI system for diabetic retinopathy detection that works with an OCT machine installed in a van, travelling to the remotest parts of the outback. The mobile AI service is able to provide earlier diagnosis for Indigenous Australians and other underserved communities, who suffer a disproportionately higher burden of disease with limited access to healthcare.

The next phase of the programme will have an oculomics function, detecting risk of cardiovascular disease and enabling earlier intervention. Other colleagues are exploring AI tools that can be used on a smart phone, using only external images of the eye to provide diagnosis in low-resource settings in Africa and Asia.

What advice would you give to your younger self?

Take up film-making as a hobby. If I hadn’t gone into ophthalmology and data science, I might have pursued a career as a film director. I am fascinated by imaging technology, both in a medical context, and in the creative arts. Also, in the same way that my team and I work with huge amounts of ophthalmic imaging data, curating it into structured datasets and optimising its use for healthcare research, the film maker needs to create a coherent narrative from a lot of raw footage, so I see interesting parallels!