AI tools detect kidney disease from eye photos

Researchers have developed AI-powered tools that can screen for chronic kidney disease using retina photos

Spotted: The Singapore Eye Research Institute (SERI) has developed two new artificial intelligence (AI) tools that use retinal images to screen for disease. The first, called RetiKid, uses retinal photos to screen for kidney disease. The second, called RetiAge, was developed in conjunction with Medi Whale, a healthcare startup in South Korea. RetiAge uses retinal photos to predict the 10-year risk of kidney disease and death.

Both tools are based on the ‘close biological relationship’ between the retina and the kidneys. This relationship means that problems with blood vessels in the retina can correlate to changes in blood vessels in the kidneys. The blood vessels in the retina can also provide clues about the ageing process and the overall health of the blood and brain.

RetiKid was trained with more than 23,000 retinal images from around 12,000 participants in Singapore and China. In a study, it had an accuracy as high as 91 per cent and has since been licensed to startup EyRIS for production and commercialisation. The tool can also be automated, for use in mass screening programmes, and could in future be connected with smartphones, enabling point-of-care diagnosis. RetiAge has been tested on more than 129,000 retinal images from around 40,000 South Korean individuals and is currently being refined further.

Both tools could offer a vital, non-invasive approach to initial diagnosis. Cheng Ching-Yu, head of SERI’s Ocular Epidemiology Research Group and Data Science Research Platform pointed out that, “The retina is a non-invasive window into one’s biological age and systemic health status, and can tell us many things about a person’s morbidity and mortality risks”. 

The SERI tools are part of a growing trend in the use of AI-powered predictive technology to aid in the identification of disease risk. At Springwise, we have covered a number of innovations in this space, including an eye test that can identify eating disorders and AI that scans x-rays for additional health conditions.