Dean Rapper
Diabetic retinopathy cost Terry Quinn his sight
This is the second article in a six-part series examining how AI is changing medical research and treatment.
Terry Quinn was just a teenager when he was diagnosed with diabetes. In a way, he rebelled against the label and frequent testing, not wanting to feel different.
His greatest fear was that one day he would have to have his foot amputated. Vision loss, another possible complication of diabetes, wasn't really on his radar. “I never thought I would lose my sight,” says Quinn, who lives in West Yorkshire.
But one day he noticed that he was bleeding in his eyes. Doctors told him he had diabetic retinopathy: damage to the blood vessels in the retina linked to diabetes. This required laser treatments and then injections.
Ultimately, the treatments were not enough to prevent the deterioration of his vision. He would hurt his shoulder walking into streetlights. He couldn't make out his son's face. And he had to give up driving.
“I felt pathetic. I felt like a shadow of a man who couldn’t do anything,” he recalls.
One thing that helped him out of his despair was the support of the Guide Dog Association for the Blind, who matched him with a black Labrador named Spencer. “He saved my life,” says Quinn, who now raises money for guide dogs.
In the UK, the NHS invites patients for a diabetes eye screening every one to two years.
U.S. guidelines state that every adult with type 2 diabetes should be tested at the time of diabetes diagnosis and then annually if there are no problems. Yet for many people this doesn't happen in practice.
“There is very clear evidence that screening prevents vision loss,” says Roomasa Channa, a retina specialist at the University of Wisconsin-Madison in the United States.
In the United States, barriers include cost, communication and convenience. Dr. Channa believes that making testing easier to access would help patients.
To screen for diabetic retinopathy, healthcare professionals take pictures of the inner back lining of the eye, known as the fundus.
Currently, manually interpreting fundus images is “a lot of repetitive work,” says Dr. Channa.
But some believe artificial intelligence (AI) could speed up the process and make it cheaper.
Diabetic retinopathy develops in fairly clear stages, meaning AI can be trained to detect it.
In some cases, the AI could decide whether a referral to an ophthalmologist is necessary or whether it could work in tandem with human image reviewers.
Getty Images
It is recommended that diabetic patients have an eye exam every year or two.
One such system was developed by Portugal-based health technology company Retmarker.
Its system identifies fundus images that might be problematic and sends them to a human expert for further investigation.
“Normally we use it more as a support tool to give information to humans so they can make a decision,” explains João Diogo Ramos, general manager of Retmarker.
He believes that fear of change limits the adoption of AI-based diagnostic tools like this.
Independent studies have suggested that systems such as Eyenuk's Retmarker Screening and EyeArt have acceptable sensitivity and specificity rates.
Sensitivity indicates how effective a test is at detecting disease, while specificity indicates how effective it is at detecting the absence of disease.
In general, very high sensitivity could be linked to more false positives. False positives generate both anxiety and expense because they lead to unnecessary visits to a specialist. In general, poor quality images can lead to false positives in AI systems.
Getty Images
AI can be trained to examine images of the fundus – the back wall of the eye
Google Health researchers examined weaknesses in an AI system they developed to detect diabetic retinopathy.
The results were very different when tested in Thailand, compared to hypothetical scenarios.
One problem is that the algorithm required pristine fundus images. It was a far cry from the realities of sometimes dirty lenses, unpredictable lighting and camera operators with different levels of training.
Researchers say they learned lessons about the importance of working with better data and consulting a wide range of people.
Google is confident enough in its model that in October the company announced it was licensing it to partners in Thailand and India. Google also said it was working with Thailand's Ministry of Public Health to assess the cost-effectiveness of the tool.
Cost is a very important aspect of new technology.
Mr. Ramos says Retmarker's service could cost around 5 euros per screening, although with variations depending on volume and location. In the United States, medical billing codes are significantly higher.
In Singapore, Daniel SW Ting and colleagues compared the costs of three screening models for diabetic retinopathy.
The most costly was the human evaluation. However, full automation was not the cheapest, as it had more false positives.
The most affordable was a hybrid model, in which the initial filtering of results was done by AI, before humans took over.
This model has now been integrated into the Singapore Health Service's national IT platform and will go live in 2025.
However, Professor Ting believes that Singapore was able to save money because it already had a strong infrastructure for diabetic retinopathy screening.
Bilal Morning
Bilal Mateen says medical AI should be available beyond rich countries
Cost-effectiveness is therefore likely to vary considerably.
Bilal Mateen, head of AI at health NGO PATH, says the data on the cost-effectiveness of AI tools aimed at preserving sight is quite strong in rich countries like the UK, or in some middle-income countries like China. But this is not the case for the rest of the world.
“With rapid advances in what AI can do, we need to worry less about whether it's possible, and more and more about whether we're building for everyone or just a select few. We need more than just data on effectiveness to make effective decisions,” insists Dr Mateen.
Dr. Channa highlights the health equity gap even in the United States, which she hopes this technology can help close. “We need to expand it to places where access to eye care is even more limited. »
She also emphasizes that older adults and people with vision problems should consult ophthalmologists, and that the convenience of AI in regularly detecting diabetic eye diseases should not distract from all other eye diseases. Other eye conditions, such as myopia and glaucoma, have proven more difficult for AI algorithms to detect.
But even with these caveats, “the technology is very interesting,” says Dr. Channa.
“I would like to see all of our diabetic patients get screened in a timely manner. And I think given the burden of diabetes, that's a really potentially interesting solution.
Back in Yorkshire, Mr Quinn certainly hopes the new technology will take off.
If AI had existed to detect his diabetic retinopathy earlier, “I would have grabbed it with both hands.”
More business technology
Source link
