A London-based AI company has created clinical decision support tools for eye-care professionals.
Visulytix retinal imaging AI analysis platform is able to read a retinal picture or OCT scan, taken from the back of a patient’s eye by a technician, in a few seconds. The system, known as Pegasus, can then detect and screen for major blinding diseases, such as glaucoma and macular degeneration.
The system is also able to detect and highlight specific features that allow it to make its suggestion. This allows the healthcare professional to work alongside the AI system, questioning, overriding or accepting its suggestion. Pegasus can also be used to triage a high volume of patients, for example, in screening programmes.
In a recent study at Massachusetts Eye and Ear Infirmary, the AI system was found to be comparable to consensus experts in the prediction of manifest glaucoma based upon analysis of optic disc images. Furthermore, another study performed by Rutgers University (USA) showed that, on average, the AI system was eight times quicker than a human at reading a retinal image.
It’s estimated that by 2020, up to 415 million people will be suffering from diabetic eye diseases, 80 million will have glaucoma and around 196 million will be suffering from age-related macular degeneration.
Visulytix anticipates that AI systems such as Pegasus will become commonplace due to reduced costs and patients being enticed to access their own records through mobile devices.