Indiana University scientists with expertise in optometry and artificial intelligence are among the first three groups of researchers to receive support from a new National Institutes of Health program to support the emerging field of “oculomics,” which uses the eye as a lens for diseases that affect the entire body.
Stephen A. Burns, professor at the IU School of Optometry, has been named principal investigator under a three-year, $4.8 million award from the NIH Venture Program Oculomics Initiative. Among the award’s co-investigators is Eleftherios Garyfallidis, associate professor of intelligent systems engineering at the IU Luddy School of Informatics, Computing and Engineering.
The project will support the development of next-generation ophthalmoscopes; instruments to observe the inside of the eye -; who can spot the early warning signs of conditions such as diabetes, heart disease, kidney disease, sickle cell anemia and Alzheimer’s disease with a simple eye scan.
“This research is about using the eye as a window into health,” Burns said, noting that the retina is the only directly observable part of the central nervous system. “We want to give healthcare providers the clearest view they can get into the body non-invasively.”
Other researchers on the project include co-principal investigator Amani Fawzi of Northwestern University and co-investigators Alfredo Dubra of Stanford University and Toco YP Chui of Mount Sinai’s New York Eye and Ear Infirmary.
Burns’ research into using the eye to detect disease dates back to the early 2000s, when he and colleagues at the IU School of Optometry pioneered the application of adaptive optics that scan laser systems to the observation of the eye. human eye. The field was originally developed by astronomers to eliminate the “twinkle” of stars; or distortions caused by the Earth’s atmosphere -; at telescopes. The optics of the eye produce similar light distortions.
Using technology developed at the school, the ophthalmoscope in Burns’ laboratory can view the back of the human eye with a resolution of two microns -; a scale small enough to show the real-time movement of red blood cells in the arteries and veins of the eye. (A single red blood cell is about eight microns wide.) Burns has used the technology to identify biomarkers for diabetes and hypertension in the walls of the eye’s blood vessels.
Project researchers from Northwestern and Mount Sinai have used similar technology to observe the cells both outside and inside these blood vessels, including detecting the crescent-shaped red blood cells found in sickle cell disease. The Stanford researchers used adaptive optics to improve observation of the eye’s photoreceptors.
With support from the NIH, the research teams will integrate their individual projects into a single device, applying state-of-the-art machine learning and AI. In addition, they will investigate the technology’s potential to detect the early signs of heart disease and Alzheimer’s disease.
There is increasing evidence for a strong retinal vascular component in Alzheimer’s disease. Currently you can see the signs with PET scans, which require large, multi-million dollar instruments. If we can see the same signs with an eye scan, it’s a lot less invasive and a lot cheaper.”
Stephen A. Burns, professor at the IU School of Optometry
Garyfallidis’ role is to develop and apply machine learning and AI methods to interpret the results of the devices. This could reduce diagnosis time from days to minutes by eliminating the need for a human to analyze the images.
In the first year of the project, the labs will tune their instruments to the same sensitivity level, said Burns, whose lab will integrate its technology with Northwestern’s instrument. Stanford will focus on similar technology integrations with the New York Eye and Ear instrument.
Next, work will shift to data validation to confirm that the new instruments’ measurements match previous versions of the technology. The researcher will also compare the new AI system’s interpretation of scans with the conclusions of human analysts to confirm accuracy.
The final year of the project involves testing the device on clinical volunteers. Much of IU’s data will come from individuals recruited through the Atwater Eye Care Center.
“Up to 80 percent of the population over the age of 60 has at least one health problem that can be detected in the eye with our technology,” Burns said.
He said the NIH selected the project because of its potential for major impact. Venture Fund initiatives emphasize “short, modest investments that can be implemented quickly, with strong potential to accelerate science.”
“Our challenge now is selectivity and specificity,” Burns said. “We need to show that we can detect the differences between conditions, to quickly and accurately interpret the signs of the different diseases we focus on.”
The goal is to advance the technology until it’s ready to make the leap from the lab to “wherever you get your annual eye exam,” he said.