By Mark Waghorn via SWNS
Artificial intelligence can predict Alzheimer's at least two years in advance, according to new research.
It analyses routine patient information such as memory and brain function, performance on cognitive tests and lifestyle factors.
The technique works by spotting hidden patterns in the data and learning who is most at risk.
A study involving more than 15,300 people found it was 92 percent accurate.
The neural network would change the way dementia is diagnosed - helping doctors detect it sooner.
Treatments would start much earlier. One of the reasons drugs have failed is the trials start after the disease has already taken hold.
Lead author Professor David Llewellyn said: "We are now able to teach computers to accurately predict who will go on to develop dementia within two years."
The breakthrough also offers hope of cutting the number of people wrongly diagnosed.
via GIPHY
Many treatable conditions have symptoms similar to dementia. Some doctors have a hard time telling them apart.
Prof Llewellyn said: "We are also excited to learn our machine learning approach was able to identify patients who may have been misdiagnosed.
"This has the potential to reduce the guesswork in clinical practice and significantly improve the diagnostic pathway, helping families access the support they need as swiftly and as accurately as possible."
The University of Exeter team tracked more than 15,300 individuals in the US who attended 30 memory centers between 2005 and 2010.
At the outset, many were experiencing problems with amnesia or absentmindedness.
Over the next decade, around one in ten (1,568) received a new diagnosis of dementia within two years of visiting the clinic.
The algorithm identified them much better than expert neurologists - matching them to patient outcomes in the database.
via GIPHY
What is more, around eight percent (130) of the diagnoses appeared to be made in error as they were subsequently reversed.
AI (artificial intelligence) recognized more than 80 percent of these cases. inconsistent diagnoses.
The study published in the journal JAMA Network Open found it can spot those who will develop dementia - and improve the accuracy of the diagnoses.
Added co-author Dr. Janice Ranson: "We know dementia is a highly feared condition.
"Embedding machine learning in memory clinics could help ensure diagnosis is far more accurate, reducing the unnecessary distress a wrong diagnosis could cause.”
Follow-up studies are being planned to evaluate assess if it can be rolled out to improve diagnosis, treatment and care.
Dr. Rosa Sancho, of Alzheimer's Research UK which funded the project, said: "Artificial intelligence has huge potential for improving early detection of the diseases that cause dementia.
"It could revolutionize the diagnosis process for people concerned about themselves or a loved one showing symptoms.
"This technique is a significant improvement over existing alternative approaches.
"It could give doctors a basis for recommending lifestyle changes and identifying people who might benefit from support or in-depth assessments."
The number of dementia cases worldwide is set to triple to 150 million in the next three decades.
Currently, the disease affects around 920,000 in the UK, a figure that will rise to two million by 2050.
There is no cure.