Columbia University researchers are using automated speech analysis to determine if an “at risk” youth will develop psychosis. The goal is early intervention.
In a very small (34 patient) study, the system differentiated — with complete accuracy — between at-risk young people who developed psychosis over a 2.5 year period, and those who did not. The computer model outperformed screening technologies such as biomarkers from neuroimaging and EEG recordings.
An algorithm rooted out “jarring disruptions” in otherwise ordinary speech. Semantic analysis measured coherence and two syntactic markers of speech complexity — sentence length and how many clauses it entailed.
Professor Gillinder Bedi believes that “If speech analyses could identify those people most likely to develop schizophrenia, this could allow for more targeted preventive treatment before the onset of psychosis, potentially delaying onset or reducing the severity of the symptoms which do develop.”