Parkinson’s is the fastest-growing neurological condition in the US, with 90,000 people now diagnosed every year alone – a 50 per cent jump since the mid-1980s. The picture is similar worldwide: a staggering 25 million people are expected to be diagnosed by 2050, more than double today's figures.
In short, it’s a big problem. But with people living longer and populations growing, such rates are not exactly unexpected. Yet what is surprising – and frankly, a bit terrifying – is how unprepared we are for this oncoming wave.
No cure. Limited treatment. Poor diagnostic tools. Heck, we don’t even really know what causes Parkinson’s.
But before you descend into a pit of neurodegenerative despair, there is hope: scientists across the globe have been busily working away to change the Parkinson's story.
In particular, we’re about to see a revolution in how to spot Parkinson’s. Armed with cutting-edge new gadgets, AI and a radically evolving understanding of how the disease manifests across the body, researchers are on the cusp of detecting Parkinson’s not years, but decades before symptoms appear in some cases.
Currently, there is no single, definitive test for Parkinson’s. Instead, as physical symptoms – such as tremors, slow movement and muscle stiffness – emerge, a doctor may diagnose a patient based on their ability to carry out tasks such as writing or speaking.
“Neurodegenerative diseases today are where cancer was 50 years ago,” says Prof Hermona Soreq, a leading researcher developing next-generation diagnostic tools. “Namely, you do the diagnosis when it's too late to treat, when all of the damaged nerve cells have already died, so you can't get them back.”
But what if there was a way to diagnose Parkinson’s before the disease rears its ugly head? To catch – and perhaps stop – it in its tracks, before brain cells are irreparably damaged.
That's no longer just a hypothetical. In fact, there’s not even just one way to do it – there are several.
AI desk accessories
Not all diagnostic breakthroughs require a blood sample. Some could sit quietly in your office.
At the University of California, Los Angeles, Prof Jun Chen’s lab claims it has developed a diagnostic pen that can detect Parkinson’s disease simply by analysing how you write.
Its soft tip is made from a unique magnetoelastic material that shifts its magnetic field in response to pressure or bending. This effect, long known in rigid metals, was only recently discovered in soft polymers by Chen’s group, opening the door to new kinds of highly sensitive, body-friendly sensors.
“Using the magnetoelastic effect in soft materials is a new working mechanism,” Chen explains. “It is able to convert tiny biomechanical pressure, such as artery vibration, into electrical signals with high fidelity.”
The pen, which uses magnetised ink, captures both on-paper and in-air hand movements. It then sends this data to a computer, where an AI model analyses patterns associated with Parkinson’s motor symptoms.
In a pilot study, the system distinguished people with Parkinson’s from healthy controls with over 96 per cent accuracy. Even better, Chen believes that at scale, the pen could be manufactured for as little as $5 (£3.70).
“We’ve already filed a patent and want to commercialise this pen,” Chen says. “In the meantime, we want to optimise it to enhance the diagnostic accuracy.”
But if writing the old-fashioned way isn’t your thing, Chen’s team has you covered. They’ve also created a smart keyboard that works on the same principle.
As people type, the keyboard picks up subtle changes in key pressure and rhythm – often too slight to notice – and feeds that data into a machine learning algorithm.
Early tests show it can identify motor abnormalities characteristic of Parkinson’s, and the team has paired it with a mobile app to support continuous, remote monitoring.
Together, these intelligent desk tools offer a glimpse of what Chen calls a “personalised, predictive, preventive and participatory” future for Parkinson’s healthcare. A future where diagnosis could be as simple as jotting down a note or typing an email.
An eye test that could detect Parkinson’s 20 years earlier
Imagine finding out you're at risk of Parkinson’s disease during a routine eye checkup, decades before symptoms appear. That’s the promise of a new, non-invasive diagnostic technique developed by Victoria Soto Linan (and her colleagues at Laval University, Canada) using a well-established eye test called electroretinography (ERG).
The eye, Soto Linan and colleagues wrote in a recent study, offers a “window to the brain.” It’s part of the central nervous system, and visual problems like blurry vision or reduced contrast sensitivity often show up long before the better-known tremors or stiffness.
Soto Linan’s team measured how the retina – the light-sensitive layer at the back of the eye – responds to flashes of light, both in mice engineered to develop Parkinson’s-like symptoms and in newly diagnosed human patients.
They discovered a distinctive ‘disease signature’ in the retinal signals, especially in women. Crucially, this weakened signal appeared in mice well before they showed behavioural signs of the disease – suggesting the same might be possible in humans.
If so, Soto Linan believes this eye test could eventually spot Parkinson’s up to 20 years before symptoms emerge.
Read more:
- This protein could play a key role in treating Parkinson's, new study finds
- Lab-grown mini brains that mimic features of Parkinson’s created for the first time
- New treatment shown to replace neurons damaged by Parkinson's disease in mice
And unlike other early diagnostic tools still stuck in the lab, this one already has a head start.
“ERG is already used in clinics – it’s used to diagnose eye diseases,” she says. “It has a big advantage of being non-invasive, too.”
Patients sit in front of a dome that flashes lights while small sensors record how the retina responds – a process that could one day be as simple as adding an extra few minutes to your annual vision test.
The team is now working to refine the test so it can be faster and even portable, with hopes of linking it to a machine learning algorithm that gives immediate results straight to your smartphone.
The research is still in its early stages, but the implications are massive. As Soto Linan puts it, “The reason why this is so important is because this tool could potentially tell you someone’s at risk up to 20 years before that point. So, imagine how much less damage there will be by then.
"Even if we don’t have a cure, the treatments we do have could be more effective and give a better quality of life in the long term.”
Tech that spots Parkinson’s through your vocal patterns
Could your voice reveal Parkinson’s before your body does? In a recent preprint study, researchers explored whether AI could detect Parkinson’s disease simply by listening to how someone speaks.
Around 90 per cent of people with Parkinson’s develop a motor speech disorder called dysarthria, which can cause hoarseness, breathiness and an irregular pitch.
These vocal changes often show up earlier than the more visible motor symptoms like tremors, making them a promising early warning sign.
The team collected 195 short voice recordings from 31 people, including 23 with Parkinson’s. They used some of these to train four different AI models to recognise the vocal patterns associated with the disease. When tested on new recordings from the same participants, the models could identify Parkinson’s with over 90 per cent accuracy.
Because these changes are subtle and develop early, the researchers suggest voice-based screening could offer a low-cost, non-invasive diagnostic tool – possibly even delivered remotely via smartphone.
A Parkinson’s blood test
In April 2025, Soreq and her colleagues – including her son, making it a true scientific family affair – published a breakthrough new study.
What they laid out in this research was remarkable: a simple, cheap blood test based on PCR technology (remember that from COVID?) that can accurately pick up Parkinson’s, years in advance of symptoms showing up.
It works by measuring the ratio between two markers of the disease that Soreq and her team uncovered in human blood.
First, in people with Parkinson’s, the researchers found unusually high levels of certain tiny molecules called transfer RNA (tRNA) fragments. These molecules seem to follow a specific repeating pattern – called a conserved sequence motif – that sets them apart.
Simultaneously, the team found the blood of people with Parkinson’s had reduced levels of tRNA from the mitochondria (the parts of your cells often called the 'powerhouses' because they generate most of the energy your body needs to function).
“So we said, okay, ‘if we have an increase in one sequence and a decline in another, we can calculate the ratio between the increase and the decrease and then we have two factors that increase our assurance a result is real’,” Soreq says.
If the ratio passes a certain threshold, it likely means a diagnosis.
According to Soreq, a typical Parkinson’s diagnosis today may cost up to $6,000 (£4,400). The two PCR tests needed for this? A mere $80 (£60).
“It’s huge. It makes a big, big difference,” she says. With any luck, the team hopes this could be widely available within a decade, offering a potential lifeline to patients around the globe.
Read more: