Orbis receives AWS Imagine grant

With the support from the Amazon Web Services grant, Orbis will expand the deployment of its artificial intelligence-assisted technology screening for diabetic retinopathy and glaucoma

Teenage boy is leaning against a white metal bar. He is wearing a bright blue t-shirt and is smiling looking at the camera.
Serrah Galos/Orbis

Orbis has been named as a winner of the 2023 Amazon Web Services (AWS) Imagine grant, supporting the eye care charity’s work to deploy artificial intelligence (AI)-assisted technology to screen for diabetic retinopathy and glaucoma in low resource areas.

The AWS Imagine grant programme provides resources to non-profit organisations looking to deploy cloud technology as a central tool to achieve their goals. AWS seeks proposals for ideas on leveraging technology in new and innovative ways to accelerate impact in local and global communities.

The Cybersight AI technology used by Orbis analyses images of the back of the eye and indicates whether the patient shows signs of glaucoma or diabetic retinopathy.

Orbis built Cybersight AI over three years using AWS cloud services, deploying the technology in partner hospitals in Rwanda and Vietnam.

With expanded support from AWS, Orbis is now implementing Cybersight AI in the AWS Africa (Cape Town) region.

Nicolas Jaccard, Orbis principal architect, telehealth and programme technology, commented: “Cybersight AI, Orbis’s state-of-the-art technology, is a proven, effective and accessible technology to screen for diabetic retinopathy and glaucoma.”

“This technology is more scalable, cheaper and faster than current models that require a human grader,” Jaccard continued. “With this grant, Orbis will expand this revolutionary tool to more areas without access to quality eye care.”

Recent Orbis studies have suggested that Cybersight AI leads to an increased uptake of referral services amongst patients, and to be equally as effective for children with diabetes as for adults.

The AWS support will add further capabilities, such as training the technology to take advantage of larger and more diverse data sets, and also adding the detection of diabetic macular oedema.

The cost of AI-enabled healthcare platforms is typically too prohibitive for providers in low-to middle-income countries, the charity noted, suggesting that Cybersight AI has the potential to become a default screening method in these communities.

Orbis explained that its democratisation of AI technology “will provide real-time, automated clinical advice at no cost to health facilities and their patients in low- to middle-income countries.”

The charity shared the experience of Jean (pictured above), a 19-year-old living with type 1 diabetes in Kigali, Rwanda, who received a screening at the Rwanda Diabetes Association clinic.

He expressed relief at learning that he was not showing signs of disease and commented: “I will continue taking care of myself and following the recommendation from my health care worker, so my eyes don’t get damaged from diabetes.”