Sask. researchers studying eyes for early signs of Alzheimer’s disease

A new study by University of Saskatchewan researchers hopes to find early diagnostic indications of Alzheimer’s through routine eye exams.

Changiz Taghibiglou is teaming up with an artificial intelligence and machine learning expert, using routine retinal scans to try detecting Alzheimer’s decades before the signs and symptoms.

“There might be a way to early detect Alzheimer’s, maybe two or even more decades before showing the signs and symptoms,” said Taghibiglou.

This could become the first non-invasive way to detect or pre-diagnose Alzheimer’s, as one common method requires drawing spinal fluid with a long needle. Taghibiglou says earlier detection could mean better patient outcomes.

“The patient can modify their lifestyle, diet, and they will slow down the progress of Alzheimer’s disease if they know they are going to get it 20 or 30 years later,” he said. “That’s the beauty and advantage of this study.”

The type of retina scan the researchers are using is called an optical coherence tomography, or OCT scan. These are done by most optometry offices when patients go in for tests.

“We consider as neuroscientists, eyes are windows of the brain,” said Taghibiglou. “Because it’s directly connected to the brain through the optic nerve, that’s why we can see many observations inside the brain by looking at the optic nerve and retina.”

Dr. Ravikrishna Nrusimhadevara says using the OCT scanning machine, doctors can see details in the eye layers with great precision.

“We can measure the blood supply of the eyeball in the rack itself, which directly correlates with the blood supply and the health of the brain tissue in the optic nerve,” said Nrusimhadevara, an ophthalmologist in Saskatoon.

“With the use of AI, we’re trying to develop an automated algorithm for the assessment of the perfusion of the blood supply within the eye and correlate it with the brain function. And hopefully, we can determine the typical characteristics that would be required to correlate positively with Alzheimer’s, then we may be able to provide a non-invasive testing and screening tool.”

Using anonymized retinal scans from patients around Prince Albert, Saskatoon and Regina, Sara Mardanisanami has a large enough dataset to work with. After pre-processing by labeling for things like sex, age and stage of the disease, she will feed that data into an artificial intelligence algorithm.

Her hope is that they find markers in the layers of the retina, or levels of deterioration, that may be present in early scans of patients who develop Alzheimer’s.

“A.I. can help to extract some features maybe that we cannot see by our eyes or we cannot, for example, detect by something,” she said. “I think A.I. can be beside the physicians and help them get more information for their prediction.”

While the researchers are focused on detecting markers for Alzheimer’s in this study, they’re hopeful that this type of work could be applied to the detection of other diseases in the future.

The study, “Screening eyes of people in Saskatchewan with non-invasive imaging technologies and artificial intelligence for early detection of Alzheimer’s Disease,” has been given a $150,000 grant from the Saskatchewan Health Research Foundation.