Eyetelligence Workstation, an artificial intelligence (AI) platform developed in Melbourne to detect signs of glaucoma, is now being used by 50 Australian optometry practices.
With an accuracy in excess of 95%, the inexpensive software analyses digital fundus images and also flags early signs of diabetic retinopathy (DR), age-related macular degeneration (AMD) and severe cataract. AI models such as this improve their accuracy with increasing input. Eyetelligence reported that a reassessment of its 200,000-image database has retrained Workstation, reducing false positives by a further 50%. The platform has received other updates since its 2020 launch, including compatibility with any white flash retinal camera and a three-stage system that grades the patient’s level of risk, said the company.
Thanks to funding of around AU$5 million from the Medical Research Future Fund, Eyetelligence has recently partnered with Centre for Eye Research Australia (CERA) and Monash University to develop an AI algorithm that will use fundus images to assess cardiovascular disease risks.