For the study, the researchers used automated natural language processing to develop speech analysis software. The software analyzed transcripts from interviews with two groups of at-risk youth — one in New York City with 34 participants and one in Los Angeles with 59 participants — for cues like tangential language, loose associations or reduced complexity of speech.
The speech analysis software was able to predict which participants would develop psychosis — a psychiatric disorder characterized by a patient’s “break” with reality — within two years with an accuracy of 83 percent in one of the groups. For the second at-risk cohort, the software was able to predict psychosis onset in the same period with 79 percent accuracy.
The researchers said their findings suggest similar “big data” approaches might prove useful for the prediction of various psychiatric disorders.
“Language and behavior are the primary sources of data for psychiatrists to diagnose and treat mental disorders,” said Cheryl Corcoran, MD, associate professor of psychiatry at the Icahn School of Medicine at Mount Sinai and first author for the study. “This technology could be applied across psychiatry, and plausibly in other fields of medicine.”
More articles on data analytics & precision medicine:
NIH seeks general public’s insights for ‘All of Us’ precision medicine program
Report: Global direct-to-consumer genetic testing market to grow 25.6% by 2022
Microsoft, UCLA researchers are using AI to make CRISPR more accurate
