How Siegfried Wagner is using AI to unlock the connection between eye health and general health

“The reason I got into ophthalmology in the first place was because I found the eye was the only part of the body that really united the heart, the brain, the immune system.” explained Dr Wagner, who is now a clinical research fellow at NIHR Biomedical Research Centre’s Artificial Intelligence Research Hub in Moorfields Eye Hospital, “You can learn so much from a single small niche of the body. I find that incredibly motivating and inspiring.”

An early epiphany

It was on a visit to an eye clinic in South Africa during a medical student placement, that Dr Wagner came to a realisation. “Every patient in that eye clinic was diagnosed with a systemic disease based on their eye condition,” he observed, “They came in with blurred vision and they went out with a diagnosis of tuberculosis, or HIV, or malnutrition, or a Vitamin A deficiency.”

Patients weren't concerned about symptoms like coughing or having ulcers in their mouth, he remembers, what motivated them to come in was their vision. “As soon as they lost their sight, they cared and they came into hospital, and that just showed me what people will do in terms of their vision,” he explains. “Even though you're not saving lives, it's amazing how much more gratifying ophthalmology work is compared to other fields. It's a real privilege because patients are so grateful.” 

Cementing key connections

It’s this ‘niche’ that has been the springboard of Dr Wagner’s whole career. “We’ve probably known for hundreds of years that the eye holds secrets about our general health, telling us about everything from blood pressure to TB, but because nowadays we have such advanced imaging techniques and sophisticated ways of analysing those images, we can pick up incredibly subtle changes. So, it’s a really exciting time to be working in this area.”

Dr Wagner is currently helping to run one of the largest studies on retinal imaging scans and Parkinson’s disease to date. He, along with a team of scientists from Moorfields Eye Hospital and the UCL Institute of Ophthalmology, led by Professor Pearse Keane have been using artificial intelligence to analyse thousands of optical coherence tomography (OCT) scans of patients’ eyes from the AlzEye database and the UK Biobank, and then using the data they collect to spot trends and connections to other conditions. 

AI was a critical tool in the study, thanks to its speed and accuracy in image analysis. “To analyse more than 150,000 eye scans, fully automated tools were necessary. Doing that manually would take, probably not even years, it would take decades,” Dr Wagner explains. Thanks to the use of AI, the team have been able to detect signs of Parkinson’s in patients' eyes up to seven years earlier than current techniques. None of which would have been possible without the collaborative funding for the AlzEye database by Fight for Sight and Alzheimer’s Research UK in 2017. 

The pros and cons of living in ‘The Data Age’

“I think there’s a common misconception that patients assume we know a lot about their general health, after all the NHS has often been their primary caregiver since birth,” Dr Wagner has observed, “but in fact, we rarely have a complete picture of a patient’s general health.” Data collection is in fact relatively new, so that just hasn’t been historically possible. 

So now, as we find ourselves in what’s been referred to as The Data Age, the team has had to be ethically stringent, after all people are wary about sharing personal details, particularly when it comes to medical records. Dr Wagner has spent the last few years working with Professor Keane, obtaining ethical approval for this unprecedented work at every stage. 

From the patient perspective

Now the concept has been proven, and the ethics adhered to, the potential for the AlzEye project is huge. “From a patient’s perspective it’s key that we can make observations in the eye in a non-invasive way.” Dr Wagner enthuses, “Blood tests and brain scans aren’t very pleasant, to say the least; for some brain scans you need to lie there for an hour remaining absolutely still in a claustrophobia-inducing space. Oculomics offers a risk free, fast, non-invasive alternative to scans and blood tests and perhaps best of all, it’s something that people are already doing! We know that people treasure their vision and often prioritise eye tests over other things, so we’re piggybacking on that. It’s low-friction and people can be screened even who feel well in themselves but might actually have underlying issues.”

Though the AlzEye study focussed on the connection between eye health and Parkinson's, Parkinson’s isn’t the only condition with early signs detectable through ocular examination. Scientists like Dr Wagner have been able to identify everything from high blood pressure, diabetes and even gum disease, to the early signs of multiple sclerosis and schizophrenia, just by looking into people’s eyes. 

How research funding has unlocked the power of AI

Dr Wagner is insistent that this work wouldn’t have been possible without vanguard funding from Fight for Sight “That priming from Fight for Sight at the start of the project was essential and gave us a clear goal to prove pilot evidence for a more substantial grant. And ultimately it was successful on so many fronts.”

“It’s been the springboard to a whole new set of ideas and collaborations across the UK and overseas.” Dr Wagner enthuses.

And AI technology holds further promise in this field; in 2019, Fight for Sight also funded research from universities using AI to analyse eye scans for the early warning signs of Birdshot Uveitis, and an AI research project funded by Fight for Sight has already identified drugs that could be repurposed to treat macular degeneration. Could it lead us to discover novel treatments for other eye diseases as well? Dr Wagner's response is unequivocal: “Absolutely, yes.”