Two novel artificial intelligence tools offer non-invasive methods to detect and predict underlying health issues using retinal images
THE eyes are not only windows to one’s soul, but also offer a glimpse into a person’s biological age and overall health status.
The Singapore Eye Research Institute (SERI) has worked with collaborators to develop two new screening tools — RetiKid and RetiAge.
RetiKid uses artificial intelligence-based deep learning algorithms to scan retinal images to detect chronic kidney disease (CKD), while RetiAge is able to predict one’s biological age.
“The kidney and eye share a close biological relationship. While blood vessels in the kidneys cannot be examined readily, blood vessels in the retina can be visualised directly using digital retinal photography.
“Therefore, examining blood vessels in the retina can provide clues to problems with the kidney’s blood vessels,” said Associate Professor Charumathi Sabanayagam, Deputy Head, Ocular Epidemiology Research Group, SERI, and Principal Investigator of the RetiKid study.
Early detection of CKD is challenging, as the condition does not always present with obvious symptoms in its initial stages. Testing a blood sample to measure serum creatinine level is the recommended screening test to detect clinically significant CKD. Despite the wide availability of serum creatinine testing, screening adherence is generally poor among populations at risk.
With an ageing population and rising prevalence of diabetes and hypertension, prevalence of CKD is expected to grow even more. Therefore, scaling up serum creatinine testing is a concern in many countries where the burden of CKD or renal failure is high, which will be more challenging in terms of manpower and logistics.
Since retinal images are acquired non-invasively, it can be used as a preliminary screening tool before testing blood for creatinine. In addition, those with a fear of needles may find the approach more bearable, potentially leading to a higher take-up rate.
“Drawing blood may not be practical in all settings. RetiKid has the potential to automate the screening process, thus enabling a larger number of patients to be screened at more sites. With results available within an hour, it can be an effective primary case-finding tool that complements existing screening strategies to detect CKD in general populations and high-risk groups, such as those with diabetes,” Assoc Prof Charumathi added.
By introducing RetiKid as a preliminary screening tool, individuals who screen positive may be more convinced to undergo further tests to confirm the diagnosis. Timely detection allows for prompt intervention and helps delay disease progression. Those who test negative can be scheduled for a follow-up screening at an interval appropriate for their health condition and risk factors.
Developed by SERI and the National University of Singapore’s School of Computing in 2019, RetiKid has been tested with more than 23,000 retinal images from close to 12,000 participants from Singapore and China. Results of the study, which showed that the RetiKid algorithm identified CKD with up to 91 per cent accuracy, were published in The Lancet Digital Health in May 2020.
Predicting lifespan Chronological age is the age based on one’s birthday, whereas biological age refers to the age of the body cells and reveals the physiological changes associated with the ageing process, which can be used to assess one’s general health status.
“Blood vessels in the retina are indicative of the health of one’s circulatory system and even the brain. By using digital technology on retinal images, we can predict a person’s biological age, and in turn, their risk of systemic diseases and lifespan,” said Professor Cheng Ching-Yu, Head, Ocular Epidemiology Research Group and Data Science Research Platform, SERI, and Principal Investigator of the RetiAge study.
Developed by SERI and South Korean healthcare start-up Medi Whale Inc. in 2021, the RetiAge algorithm was developed using more than 129,000 retina photos from over 40,000 participants from South Korea to predict the probability of a person having an ‘older’ retina. Its ability to predict a person’s 10-year risk of systemic disease and death was further evaluated among some 56,000 participants in the UK Biobank.
Results showed that those with the ‘oldest’ retinas had twice the risk of 10-year all-cause mortality, triple the risk of cardiovascular disease mortality and 1.7 times the risk of cancer mortality than people with the ‘youngest’ retinas, even though the two groups have the same chronological age.
Compared to the standard method of measuring biological age via DNA examination, RetiAge is a non-invasive and relatively time-saving tool that can be easily adopted in clinics. Once a person is found to have ‘older’ retinas, he can take steps to improve his health, such as exercising regularly, maintaining a balanced diet and avoiding smoking.
RetiKid has been licensed to local health tech start-up EyRIS for regulatory clearance and commercialisation to benefit more patients. With the advancement of imaging, cloud computing and mobile technologies, RetiKid has the potential to be integrated into smartphones in the future.
Meanwhile, RetiAge researchers are refining the algorithm to optimise its predictive performance in the local population. They are also studying whether the tool can be used to predict other age-related diseases.
Both RetiKid and RetiAge can be integrated with the Singapore Eye Lesion Analyser Plus (SELENA+), a retinal image-based deep learning system also developed by SERI and licensed to EyRIS, which is currently available at polyclinics for screening of diabetic eye diseases, glaucoma and age-related macular degeneration. This means that patients can be screened for more diseases with a single retinal image.
As part of RetiKid’s validation phase, SERI and the National Kidney Foundation have partnered for a community outreach programme to recruit 1,200 participants at high risk of developing CKD.
These include family members of patients who have been diagnosed with CKD or kidney failure, patients with diabetes and hypertension who do not go for regular follow-ups and Malay adults who are at high risk for CKD.
The participants will be screened using RetiKid as one of the modalities. The project kicked off in February 2022 and will continue till January 2024.
Chronic kidney disease (CKD) in Singapore
In Singapore, 5.5 per cent of adults aged above 24 years and 12.5 per cent of adults aged 40 to 80 years have CKD. Globally, Singapore ranks fourth for prevalence of CKD, and first for diabetes-induced kidney failure — two in three cases of kidney failure are caused by diabetes. Many patients are medically unfit to go through a transplant surgery and rely on dialysis for survival.