Artificial intelligence is being deployed in research to help the 1.3 million Australians with diabetes avoid sight threatening complications.
Professor Mingguang He is leading a team at the Centre for Eye Research Australia, in partnership with Australian company Eyetelligence and Monash University, to develop and deploy AI technology to improve testing for eye conditions and cardiovascular disease.
The project has received major backing from the Medical Research Future Fund (MRFF) which has provided the team almost $5 million over the next three years to test the technology in real-world situations with consumers and clinical experts from a range of disciplines. St Vincent’s cardiology and other health industry partners are also involved in the project.
Professor He says the project will see the AI-enabled technology commercialised in Australia and beyond, to benefit patients globally.
The AI system will screen patients for eye and heart disease as well as other diseases like stroke.
It will enable primary health professionals to detect conditions earlier and improve access to screening services in regional and remote areas.
How AI screening works
The team aims to improve accuracy by testing the system in primary care settings such as GP clinics, Indigenous health services, eye care clinics, and endocrinology and cardiology services.
After taking a photo of the back of the patient’s eye, the non-invasive AI system scans for signs of diseases and instantly prints a report on whether they need specialist assessment and treatment.
The AI tool’s algorithm has been developed over five years using more than 200,000 images of the back of the eye and has been found to be highly accurate.
It can also identify those at higher risk of a heart attack or stroke as changes in the retina and retinal vessels can predict the risk of cardiovascular disease.
The new technology is being trialled with more than 1000 patients, including Indigenous people in remote communities, older Australians and people with diabetes. The clinical aspect was put on hold in 2020 due to COVID-19 but is set to resume in 2021.
The Research will also compare the AI tool against current telehealth models and measure accuracy, cost-effectiveness, ease of use and patient and clinician acceptance.
To help health professionals prepare for the technology revolution Professor He is leading a series of AI seminars throughout 2021.
He says that ultimately this new technology could help 1.3 million Australians with diabetes reduce their risk of sight-threatening complications while also tackling cardiovascular disease.
Professor He says most of the technical development on the system is completed and has received Therapeutic Goods Administration approval.
“In the near future, we want it to make an impact on these health conditions, improve diagnostic accuracy in a standardised way and ensure these tools are well received in the community,” Professor He says.
“This effective model enables screening programs in remote and regional areas that are missing out on eye care services.
“It could improve the health of many Australians and save health dollars by diagnosing conditions earlier.”
Ultimately, he hopes to enhance eye condition diagnosis and make it more efficient and targeted.
“My ultimate goal is to use technology to transform the way we practise medicine,” he says.
Professor Mingguang He leads CERA’s ophthalmic epidemiology research and has a particular interest in twin studies, randomised clinical trials and imaging technology.
He is a global expert in vision-related clinical and epidemiologic research. He has led some important epidemiological studies and clinical trials, including the first population-based study on myopia in China,
He has received more than $10 million in major research grants in Australia, including the prestigious NHMRC investigator Grant and Medical Research Future Fund for his research on artificial intelligence.
He is also the Director of the World Health Collaborating Centre for the Prevention of Blindness based at CERA.
You can read more about Professor He’s research on the CERA website