Corpuscle’s glaucoma classification tool touted for national rollout
Glaucoma-induced blindness is the leading cause of irreversible blindness in the UK, yet many patients are not able to access support for their condition, but a newly developed tool hopes to change this.
Patients with glaucoma with visual impairment are eligible for financial and social support from local authorities, but only when they are in possession of a certificate of visual impairment (CVI), for which registration depends on subjective assessment and many eligible patients miss out as a result.
In a bid to improve classification and save time, recent graduate Dr Arun Thirunavukarasu (m.2017) led a research project at Cambridge Eye Research Centre to develop a Glaucoma Field Defect Classifier (GFDC) for non-specialist clinicians and researchers to identify eligible patients instantly and without a consultant ophthalmologist.
According to their latest research paper published in the British Journal of Ophthalmology, Arun and his team have developed an accurate, interpretable, and safe algorithm to automatically categorise severity of disease and eligibility for sight impairment registration. The team have been praised for making the technology freely accessible online, which means it could be incorporated into glaucoma clinics around the country and beyond.
“It has been a real team effort to get to this point," says Arun. "With special thanks to my fellow Corpuscle Dr Federico Lattuada who developed the computer vision model that the GFDC uses. I would also like to extend my thanks to the large group of students who extracted data from thousands of records to make the deployment of GFDC possible.”
Arun is currently working as an Academic Foundation Doctor and Clinical Research Fellow at Oxford University Hospitals NHS Foundation Trust.
Read more about the project here