Leveraging Artificial Intelligence to Optimize Retinal Detachment Management: A retinal Shift Analysis Approach

Brief plain language background 

Surgery to fix retinal detachment associated with a tear —a condition where the retina peels away from the back of the eye—has evolved a lot over time. Today the commonest way involves removing the gel inside the eye (pars plana vitrectomy) with a gas bubble. However, even when the retina is successfully reattached, there’s still a risk that it doesn’t line up exactly the way it was before. This can affect vision, cause distortion and even interfere with the brain. Some studies report that 6 out of 100 patients have this problem, while others say it is as many as 72 out of 100 times. Because of this, there's a growing need for better ways to monitor the eye during and after surgery and to make sure the retina stays in the right position. Helping patients avoid long-term problems. 

What problem/knowledge gap does it help address 

A significant number experience retinal misalignment, complaining of poor vision, distortion and at times significant issues with their vision after the retinal detachment surgery. We are still unsure how to correctly position patients after surgery. We do not know if the opposite/fellow/other eye (without the detachment) can be used to reduce the incorrect position (misalignment) after surgery? 

Aim of the project 

To develops an AI tool to guide post- surgery recovery after retinal re-attachment. It will do this by analysing vascular patterns, comparing with past histories and the companion eye to make a best guess of whether there is a misalignment, and if so how to adjust. 

Potential impact on people with sight loss 

This AI tool aims to enhance retinal re-attachment surgery precision, reducing complications and improving vision outcomes for patients with sight loss. It aims to provide real-time guidance on whether realignment is required, and thus ensure better patient outcomes with less distortion, poor vision and issues with the brain adapting the misalignment. The project aims to take an early technology through to phase I trials, and if successful garner much larger funding in the future.