Almost nothing frightens us as much as mental illness. If only we had a way of knowing who was likely to succumb to it. We could anticipate a psychotic break and, optimally, deliver an effective therapy to treat or manage it.

IBM physicist Guillermo Cecchi and his colleagues in computer science and psychiatry have developed an algorithm that infers a schizophrenic break from an individual’s language patterns. In this episode of Inside IBM Research, he talks about a multidisciplinary study that will help psychiatrists anticipate serious mental illness — and help society ease human suffering.

Check back in January 2016 for a complete transcript.

Related links

Guillermo Cecchi
http://researcher.watson.ibm.com/researcher/view.php?person=us-gcecchi

Inside IBM Research
http://researcher.watson.ibm.com/researcher/view_group.php?id=5665

Why Are Computers So Accurate at Predicting Schizophrenia? [Forbes]
http://www.forbes.com/sites/michaelthomsen/2015/09/20/why-are-computers-so-accurate-at-predicting-schizophrenia/

Computers Can Predict Schizophrenia Based on How a Person Talks [The Atlantic]
http://www.theatlantic.com/technology/archive/2015/08/speech-analysis-schizophrenia-algorithm/402265/

Look Who Is Getting Into Mental Health Research [NIH]
http://www.nimh.nih.gov/about/director/2015/look-who-is-getting-into-mental-health-research.shtml

Word Analysis Proves Effective Pre-Psychosis Diagnosis [Building a Smarter Planet]
http://asmarterplanet.com/blog/2015/08/word-analysis-proves-effective-pre-psychosis-diagnosis.html



Using An Algorithm To Predict Schizophrenia

This is the official Soundcloud account for IBM Research. It is managed by Chelsea Dubner and Chris Nay and follows IBM social media guidelines.