A new study has found that machine-learning artificial intelligence (AI) can successfully detect and predict which individuals would go on to develop psychosis with an accuracy of 93%. The study, which was published in published in npj Schizophrenia, showed machine learning says that frequent ‘sound words’ (words associated with sound) is one clue to the later emergence of psychosis. The researchers also developed a new machine-learning method to quantify the semantics in people’s everyday conversational language.
The researchers investigated potential linguistic indicators in 40 participants of the North American Prodrome Longitudinal Study (NAPLS). NAPLS works with youth in the San Francisco Bay area that believe they may be at-risk for psychosis. The onset of Psychotic disorders and Schizophrenia typically occurs in peoples’ early 20s, with the warning signs (prodromal syndrome) beginning around age 17.
Researchers also found that speaking with low semantic density (or more vagueness) was also a sign for onset of psychosis.
“Trying to hear these subtleties in conversations with people is like trying to see microscopic germs with your eyes,” says Neguine Rezaii, first author of the paper. “The automated technique we’ve developed is a really sensitive tool to detect these hidden patterns. It’s like a microscope for warning signs of psychosis.”
The study shows that AI and machine learning can be useful in early detection and prediction of psychosis, for which currently there is not accurate measure. Machine learning learns from experience and from patterns. Because of that, machine learning can spot patterns in people everyday conversational language that may slip past doctors, even those who have undergone training to spot and treat those at risk for psychosis.
“The results point to a larger project in which automated analyses of language are used to forecast a broad range of mental disorders well in advance of their emergence.”
Aside from detection, the study also shows an glance inside the thought processes affected in the mind of someone with the onset of psychosis. The results were consistent with previous works suggesting that patients with psychosis have impairments in using words to generate higher order meaning. Further research could investigate more ways in which AI can detect and analyze mental illness, and offer insight into understanding the mechanism of which these illnesses are caused.