Research published in npj Parkinson’s Disease demonstrates that brain activity patterns captured through magnetoencephalography can reliably distinguish people with Parkinson’s from healthy controls
London, UK, 11 May 2026 — MYndspan, the first company to bring clinical-grade magnetoencephalography (MEG) brain scanning directly to consumers, today announced the publication of new research in npj Parkinson’s Disease, demonstrating that MEG can accurately identify individual cases of Parkinson’s disease from resting-state brain recordings.
The study, led by MYndspan scientists Dr. Gillian Roberts and Samuel Hardy, alongside co-authors Dr. Yali Pan and Dr. Robert Chen and MYndspan’s Chief Scientific Officer Dr. Benjamin Dunkley, used MEG data from 199 participants to investigate whether patterns of neural activity could reliably distinguish people with Parkinson’s from healthy controls. Participants were drawn from two cohorts, the Quebec Parkinson Network and the PREVENT-AD Research Group, accessed through the Open MEG Archive (OMEGA) repository. The model achieved strong classification performance (AUC-ROC of 0.87), successfully distinguishing people with Parkinson’s disease from healthy controls within the study dataset.
Parkinson’s disease is the second most common neurodegenerative condition worldwide, with the number of people living with it projected to reach 25 million by 2050. Despite decades of research, diagnosis continues to rely primarily on the appearance of motor symptoms, by which point the disease is typically well-established. Identifying the condition earlier, before those symptoms emerge, is one of the most important unsolved problems in neurology. The brain communicates using rhythmic patterns of electrical activity. In people with Parkinson’s disease, these patterns can shift toward slower activity, a phenomenon often referred to as “neural slowing.”
“Parkinson’s disease affects millions of people, and for too long the path to diagnosis has depended on waiting for symptoms that appear well after the disease has taken hold,” said Caitlin Baltzer, Founder and CEO of MYndspan. “This research is evidence that MEG can see what other tools miss, and that a short, non-invasive brain scan can produce clinically meaningful information. That is the foundation everything we are building at MYndspan sits on.”
Rather than a single brain signal, the study identified four distinct and independent patterns of brain activity, each one separately contributing to accurate detection of Parkinson’s. This matters because each pattern likely reflects a different underlying process in the disease, which may help explain why patients experience it so differently. Over time, that same level of detail could support patient stratification, helping clinicians match individuals to the therapies most likely to work for them and moving the field closer to personalised treatment, such as specific types of pharmacotherapy or deep brain stimulation. This is only possible because MEG captures both where in the brain activity is changing and when, with millimetre and millisecond precision.
Critically, the model is also interpretable and not a black box: researchers can see exactly which brain changes are driving each result. And all of this comes from a five-minute resting scan, with no tasks, no injections, no radiation. The findings point toward two meaningful near-term applications: earlier detection before motor symptoms become obvious, and more objective, brain-based diagnosis alongside clinical observation. Both represent a significant shift from how Parkinson’s is identified and tracked today.
“What this study demonstrates is that MEG can detect the neurophysiological fingerprint of Parkinson’s disease at the level of the individual, which is exactly what is needed for it to play a role in clinical practice,” said Dr. Benjamin Dunkley, Chief Scientific Officer of MYndspan. “The fact that we achieved this using resting-state recordings processed through our analysis pipeline is significant. It points directly to the kind of scalable, non-invasive approach that could one day support earlier detection.”
The findings build on MYndspan’s 2024 publication in Imaging Neuroscience, which demonstrated MEG’s capacity to predict brain age across the adult lifespan and identify markers of accelerated ageing linked to injury or disease. Together, the two studies establish a growing body of peer-reviewed evidence for MEG-derived biomarkers as clinically meaningful indicators of neurological change.
The study was conducted using data from the Open MEG Archive (OMEGA) repository and included collaboration with the PREVENT-AD Research Group and the Quebec Parkinson Network. Clinicians, MEG labs, and individuals interested in submitting existing recordings for analysis can learn more at myndspan.com/upload.
Learn more about MYndspan and its research at www.myndspan.com.