docet

Key Points

• EyePACS is a diabetic retinopathy screening programme that has been widely adopted in the US primary care sector and its use has increased patient compliance with screening.
• Using the capability of telemedicine, the EyePACs system forwards the images to an ophthalmologist for assessment and eliminates the need for the patient to visit the ophthalmologist for the outcome.
• The additional use of AI to automatically assess images can reduce the time taken for an image to be assessed and, in locations where ophthalmologist numbers are limited, a high volume of people can be assessed so that those at risk of visual loss can be identified and treated efficiently.
• The California Healthcare Foundation sponsored an industry competition where teams of developers from around the world competed to create the ideal diabetic retinopathy screening algorithm. This competitive drive led to hundreds of algorithms being developed in just six months and AI could now be harnessed to grade diabetic retinopathy with speed, accuracy and consistency.
• Google went on to use the huge databank of EyePACs images and developed further even more advanced AI algorithms. Using insights from research carried out with medical anthropologists, a more detailed grading system was created that could direct the optometrist and encourage better engagement with patients.