The team involved 80 80-year-old participants, 50% of whom were diagnosed with the disease. Before the onset of symptoms, the subjects had been subjected to a language test, during which they were asked to describe a graphic representation
–
London – Predict the possibility that it will arise Alzheimer’s disease years in advance simply by looking at a subject’s writing pattern. A study, published in the journal, aims at this goal ClinicalMedicine, conducted by experts from the Thomas J. Watson Research Center, who developed and trained artificial intelligence to gather clues to the change in language before the onset of neurological disorders and the onset of symptoms.
“The prediction of the onset of Alzheimer’s is only the first in a series of studies aimed at using artificial intelligence to evaluate the possibility of neurological disorders – he says. Ajay Royyuru del Thomas J. Watson Research Center – which can bring out unique linguistic patterns, useful as early warning indicators “.
The team involved 80 participants aged 80, 50 percent of whom were diagnosed with Alzheimer’s disease. Before the onset of symptoms, the subjects had been subjected to a language test, during which they were asked to describe a graphic representation. “We analyzed the use of words by the participants – explains the expert – through an artificial intelligence program capable of detecting subtle predictive variations of Alzheimer’s disease. with a 75 percent accuracy rate. This is the first concrete evidence regarding the possibility that language can predict the onset of neurological disorders “.
The scientist explains that, when there is the possibility of delaying or arresting the disease, it will be essential to be able to diagnose the disease in advance, conducting simple tests that can identify individuals potentially at risk.
“This relationship is opening new horizons – he concludes Michael Weiner, Alzheimer’s expert at the University of California at San Francisco – and it is truly one of a kind, a milestone ”.
“For the next steps – concludes Royyuru – we hope to extend the work to other neurological disorders, in order to train artificial intelligence to recognize the subtle changes in language that could be an indication of high risk for neurodegenerative conditions “.
–
