A trio of researchers have developed an experimental machine learning method that allows AI to listen for the early whispers of psychotic break that humans can't hear. The team, consisting of Neguine Rezaii of Harvard Medical School and Emory School of Medicine, and Elaine Walker and Philipp Wolff from Emory University's Department of Psychology, set out to see if there was any way to use language as an indicator of impending latent onset psychosis. They developed a machine learning method that looks for specific indicators long thought associated with psychosis, especially schizophrenia. The team then spent two years observing study volunteers, a significant portion of whom ended up demonstrating psychotic break (the first experience of a fully psychotic episode). The results of the study were incredible. The team not only determined their tool could experimentally predict psychotic break with higher-than-human accuracy, but also discovered a new indicator of impending psychotic break.

People convey meaning by what they say as well as how they say it: Tone, word choice and the length of a phrase are all crucial cues to understanding what's going on in someone's mind. When a psychiatrist or psychologist examines a person, they listen for these signals to get a sense of their wellbeing, drawing on past experience to guide their judgment. Researchers are now applying that same approach, with the help of machine learning, to diagnose people with mental disorders. In 2015, a team of researchers developed an AI model that correctly predicted which members of a group of young people would develop psychosis--a major feature of schizophrenia--by analyzing transcripts of their speech. This model focused on tell-tale verbal tics of psychosis: short sentences, confusing, frequent use of words like "this," "that," and "a," as well as a muddled sense of meaning from one sentence to the next.