Artificial intelligence can now reveal with incredible accuracy which individuals may develop dementia, new research has found.
AI now has a 92 percent accuracy rating for predicting which memory clinic attendees will have dementia within two years, according to the study published Thursday in the journal JAMA Network Open.
The findings are based on data from over 15,300 US patients. Authors say the algorithmic accuracy of AI predictions may be able to reduce the amount of false dementia diagnoses — and possibly help doctors intervene earlier.
“We know that dementia is a highly feared condition. Embedding machine learning in memory clinics could help ensure diagnosis is far more accurate, reducing the unnecessary distress that a wrong diagnosis could cause,” said study co-author and University of Exeter research fellow Dr. Janice Ranson in a press release.
Indeed, in addition to bolstering trust in the abilities of computer prediction, the finding offers evidence that machines can help identifying “patients who may have been misdiagnosed” said Professor David Llewellyn, who oversaw the study. “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.”
Misdiagnosis is no small problem when it comes to dementia: Study researchers determined that approximately eight percent of dementia diagnoses in their large data sample were made in error and subsequently reversed. Machine learning was able to accurately identify over 80 percent of these inaccurate diagnoses.
Scientists are hopeful the findings may mean not only more accurate dementia diagnoses moving forward, but also earlier ones.
“Artificial intelligence has huge potential for improving early detection of the diseases that cause dementia and could revolutionize the diagnosis process for people concerned about themselves or a loved one showing symptoms,” commented Dr. Rosa Sancho, the head of research at Alzheimer’s Research UK.”
Sancho added that the technology “could give doctors a basis for recommending life-style changes and identifying people who might benefit from support or in-depth assessments.”