Predicting irreversible loss in glaucoma

Glaucoma is a leading cause of permanent vision loss in Australia. By 2040, around 112 million people globally are expected to have glaucoma. Despite treatment, within 20 years of diagnosis, one in three people lose vision in one eye, and one in seven lose vision in both eyes.

A major reason is the lack of sensitive clinical tools to detect disease progression. Current clinical tools, like visual field tests, need years of testing before reliably detecting disease progression. By then, substantial vision loss has already occurred. This represents a significant missed opportunity for early treatment escalation.

To help prevent permanent vision loss from glaucoma, we need a sensitive and functionally relevant way to measure disease progression. This project explores using novel artificial intelligence (AI) approaches using state-of-the-art retinal imaging to provide an objective measure of disease progression.

This research uses widefield, 3D and high-resolution retinal imaging collected as part of a study at the Centre for Eye Research Australia (CERA). Participants also underwent comprehensive visual sensitivity testing at precise retinal locations using an innovative functional assessment method.

This project will apply MDAP’s expertise with complex AI-based approaches to predict visual sensitivity in rich imaging data. Such AI-based predictions could provide vital, sensitive and functionally relevant measure of disease for the clinical management of glaucoma. This measure could also accelerate treatment discovery by improving how promising new therapies are evaluated in clinical trials.