Scientific research in the fields of neurobiology and neuropathology is making waves. Researchers now have a method for predicting dementia that is more than 80% accurate, and able to preempt a dementia diagnosis by up to nine years.
The predictive test is the first of its kind. It works by analysing functional MRI scans and looking for changes in the brain’s “default mode network”, or DMN. The researchers, based at the Centre for Preventive Neurology in London and led by Professor Charles Marshall, used fMRI scans from more than 1,000 volunteers.
New predictive test could preempt dementia diagnoses by almost a decade
Researchers from the London-based Centre of Preventive Neurology have come up with a method for predicting dementia. It’s the first of its kind, and is over 80% accurate.
Moreover, it has the ability to predict dementia up to nine years before a formal diagnosis.
Previous methods of predicting dementia include running memory tests and measuring brain shrinkage. But both of these have their inaccuracies and pitfalls.
This new test bulk-analyses fMRI scans in an effort to detect changes in the brain’s default mode network (DMN). According to the QMUL news team’s writeup of the breakthrough, the DMN connects regions of the brain to perform specific cognitive functions. It’s therefore the first neural network Alzheimer’s disease (and other forms of dementia) affects.
That’s why it’s a good place to look to see if someone is likely to develop dementia in the near future.
The researchers used fMRI scans from over 1,100 volunteers
UK Biobank, a biomedical database, connected the researchers to more than 1,100 volunteers, who allowed them to use their brain scans. Using the scans, the researchers from the Centre for Preventive Neurology estimated the effective connectivity between ten key regions of the brain – the DMN.
They then assigned each volunteer with a probability value. This represented the likelihood that they had dementia. They based the probability value on how similar their effective connectivity pattern was to patterns that indicate dementia, or a control pattern.
They compared their predictions to patients’ medical data, which the UK Biobank provided. And their findings were stunningly accurate. The model they had developed predicted the onset of dementia several years in advance of an official diagnosis – in some cases up to nine years ahead of time.
And the test was more than 80% accurate. But that’s not all. In cases where volunteers had gone on to develop dementia, the model had predicted when the disease would strike, to within two years.
What does this mean for the future of dementia diagnoses and treatment?
QMUL quotes Charles Marshall, who led the research team, as saying such predictive tests will be “vital” for developing treatments. Symptoms of dementia are the result of “irreversible loss of brain cells”, he adds. Halting that loss before it’s too late could prove transformative for would-be patients.
Alzheimer’s disease, which is a specific brain disease rather than a general term for a decline in mental ability, is caused by specific proteins in the brain. Scientists are getting better at detecting those proteins. However, Marshall goes on, “many people live for decades with these proteins in their brain without developing symptoms of dementia”.
He hopes his team’s new test will be able to identify whether people should start receiving treatments before symptoms appear.
The team’s research paper is available to read online on the Nature Mental Health website.
Charles Marshall is a clinical senior lecturer in dementia at the Preventive Neurology Unit, part of Queen Mary, University of London. He’s also the clinical lead for neurology at Barts Health NHS Trust, where he runs a cognitive disorders’ clinic. Finally, he works as a consultant in the East London Foundation Trust Memory Clinic, in Tower Hamlets.
