The problem: There are over 4 million people in the UK with diabetes, all at risk of developing diabetic retinopathy and losing their sight. People with diabetes are called for an annual eye screening appointment when a photograph of the back of their eye is assessed manually for sign of disease. However, this is subject to potential error which we need to eradicate.
Our solution: to develop powerful new image analysis tools for the automated assessment of the retinal scans. The tools will also be able to predict the progression of the disease. To get to this stage we need to undertake a process of artificial intelligence, mathematical modelling, statistical analysis, testing and validation to ensure that the software is working as hoped. The work will be carried out at the new image analysis labs at the Clinical Eye Research Centre at St Paul’s and next door in the Department of Eye & Vision Science, using real data from thousands of patients uniquely available in Liverpool as of over 20 year research in diabetes.
Wider benefits: the technology can also be used to detect and diagnose other eye diseases such as age-related macular degeneration (AMD), glaucoma and keratoconus. This has the potential to help millions of people – AMD affects more than 600,000 people in the UK alone.
Cost: £510,000 over four years
Why this is vital:
Diabetes is the next worldwide pandemic disease with projected huge increases in numbers, especially in the developing world. Losing sight is the biggest fear of people living with diabetes, but can be prevented if the earliest changes to the retina are spotted before vision is affected. We urgently need to bring new computing, artificial intelligence and imaging technologies to bear to solve this.