Neural networks have learned to diagnose Alzheimer's disease in the early stages by voice.

The experiment involved 206 people, in 114 of them, doctors revealed mild cognitive impairment. Volunteers without training were asked to describe works of art. Their speech was recorded and given to the analysis of artificial intelligence. He assessed the motor control of speech, the density of the speaker's ideas, and the grammatical complexity of statements. Diagnosis of one patient took ten minutes. Traditional neuropsychological tests usually last several hours. In most cases, artificial intelligence confirmed the diagnoses made by doctors. The use of neural networks will make life easier for patients and the selection of treatment.