Science and Research
Ophthalmic neuroscience research
Our multidisciplinary team combines expertise in ophthalmology, neuroscience, biomedical engineering, data science and machine learning. We develop novel retinal imaging technologies and analytical methods to better understand and detect neurological and retinal diseases.
Overview
The retina offers a unique window into brain health and systemic disease.
The Ophthalmic Neuroscience Unit are pioneering hyperspectral retinal imaging techniques – an advanced, non-invasive method of capturing detailed spectral signatures from the eye – to reveal subtle changes not visible on standard imaging.
Our research has contributed to the discovery of imaging biomarkers in Alzheimer’s disease, diabetic retinopathy, age-related macular degeneration (AMD), ocular tumours and glaucoma. These discoveries are driven by deep learning, advanced image processing and the largest known datasets of hyperspectral retinal images.
A key focus of our team is translational impact. We are not only building the underlying science but also designing regulatory-grade hyperspectral cameras to be deployed in clinical care.
Why this research is important
Retinal imaging is already central to eye care. However, many changes that happen early in disease remain invisible to current imaging tools. In particular, diseases affecting the brain (neurodegenerative) and blood vessels (vascular). Our work explores hyperspectral imaging can fill this gap – enabling earlier detection of blinding and brain-related diseases.
We have shown that retinal hyperspectral imaging could potentially be a non-invasive way to detect changes linked to Alzheimer’s disease. This offers an alternative to more invasive tests such as PET scans or cerebrospinal fluid analysis. As new treatments for Alzheimer’s and other brain diseases become available, retinal imaging may become a key screening tool to identify who might benefit most.
In addition to neurodegeneration, our work now focuses heavily on eye cancers (ocular oncology), glaucoma, diabetic eye disease and AMD. We are developing disease-specific ‘spectral biomarkers’ (unique light patterns associated with each disease) and AI-based diagnostic tools. These tools are trained using a large dataset, which helps create flexible, scalable solutions for use in clinics.
Key research questions
- What are the earliest detectable biomarkers of retinal and neurodegenerative diseases in hyperspectral data?
- Can we develop a foundation model from large-scale hyperspectral datasets to enable multi-disease detection from a single scan?
- How can hyperspectral imaging support disease monitoring, treatment response evaluation, and risk stratification in both eye and brain disease?