Prediction of psychotic onset with AI language analysis

January 24, 2018, IBM
The figure plots the average and standard deviation of the values of the two main factors that provide for conversion (i.e. transition from at-risk to onset) prediction. There are four groups: controls (CTR), patients at-risk (clinical high risk) who do not transition to psychosis (CHR-), patients who do eventually transition (CHR+), and patients who have already transitioned to psychosis (FEP, i.e. first episode). The horizontal axis is related to the maximal value of semantic discourse coherence; observe that the control population have higher values. The vertical axis corresponds to the variance of the coherence; in this case the patients who eventually transition to psychosis (CHR+) show a higher value, meaning that they oscillate between being highly incoherent to being highly coherent. Those who have already converted (FEP), on the other hand, have low maximal coherence, and low variance; i.e. the stay always incoherent. Credit: IBM

Psychiatrists characterize schizophrenia, a mental condition with devastating effects on those who suffer it, by a set of intuitively understandable concepts including "poverty of speech" and "flight of ideas." These concepts, however, are subjective in the sense that their quantification depends significantly on the particular training and ultimate judgment of individual psychiatrists. The evaluation of a patient is not only subject to this uncertainty, but also to the availability of these highly trained professionals, and to the clinical facilities where evaluation is usually performed.

The IBM Research team for Computational Psychiatry and Neuroimaging has been developing methods intended to mitigate these limitations using Artificial Intelligence (AI) to provide a consistent (i.e. evaluator-independent) neuropsychiatric evaluation from samples, which could be delivered remotely. In collaboration with academic researchers, we published results in 2015 showing that it is possible to use AI to model differences in between high-risk who subsequently developed and those who did not. Those results were obtained by analyzing baseline interviews in which patients talk about themselves for close to one hour, using the AI method Natural Language Processing (NLP) to quantify the notion of "poverty of speech" as syntactic complexity, and "flight of ideas" as semantic coherence. While successful, this approach had several limitations: only one cohort, from a single location, and using a single evaluation protocol.

In a recent paper, Prediction of psychosis across protocols and risk cohorts using automated language analysis, published in the top-ranked psychiatry journal, World Psychiatry, we took steps toward overcoming these limitations. Studying a different and larger patient cohort, from a different location and a very different evaluation protocol (patients being asked about how they understood a story they read), we built a retrospective model of patient speech patterns that would have predicted with 83 percent accuracy whether a patient subsequently developed psychosis. Significantly, this model also would have predicted with 79 percent accuracy whether patients in the earlier cohort would have developed psychosis. That is, the automated, AI-enabled analysis of the speech patterns of these at-risk patients may have the potential to be highly predictive of the future onset of psychosis in a way that bridges differences in recruitment criteria between institutions, and the specifics of the probing protocol.

We believe this is a significant step towards the goal of developing a tool for mental health practitioners, caregivers and patients themselves with a tool that can expand and multiply the reach of neuropsychiatric assessment outside of the clinic. Stay tuned to hear more about the work we are doing in several other such as depression, Parkinson's and Alzheimer's diseases and chronic pain, among others.

Credit: IBM

Explore further: Speech analysis software predicted psychosis in at-risk patients with up to 83 percent accuracy

More information: Cheryl M. Corcoran et al. Prediction of psychosis across protocols and risk cohorts using automated language analysis, World Psychiatry (2018). DOI: 10.1002/wps.20491

Related Stories

Speech analysis software predicted psychosis in at-risk patients with up to 83 percent accuracy

January 22, 2018
Computer-based analyses of speech transcripts obtained from interviews with at-risk youths were able to predict which youths would later develop psy-chosis within two years, with an accuracy of up to 83 percent. In two independent ...

Who will develop psychosis? Automated speech analysis may have the answer

August 26, 2015
An automated speech analysis program correctly differentiated between at-risk young people who developed psychosis over a two-and-a-half year period and those who did not. In a proof-of-principle study, researchers at Columbia ...

Early detection method hopes to prevent psychosis

October 14, 2016
Mental health researchers have made a promising breakthrough in the early detection of the risk of psychosis, with the eventual hope that patients could be given appropriate treatments earlier to prevent psychotic episodes ...

An ingredient in cannabis may be useful for treating psychosis – new study

December 15, 2017
Psychiatric patients treated with a substance found in cannabis, cannabidiol, showed a significant reduction in psychotic symptoms and were also more likely to be rated as "improved" by their psychiatrist, our latest study ...

Cancer probability documentation lacking for some at high risk

December 21, 2017
(HealthDay)—For patients with high-risk indeterminate pulmonary nodes (IPNs), physicians rarely provide quantitative documentation of cancer probability, according to a study published online Dec. 20 in JAMA Surgery.

Autism traits increase thoughts of suicide in people with psychosis

December 14, 2017
People with autism traits who have psychosis are at a greater risk of depression and thoughts of suicide, new research has found.

Recommended for you

People with family history of alcoholism release more dopamine in expectation of alcohol

May 23, 2018
People with a family history of alcohol use disorder (AUD) release more dopamine in the brain's main reward center in response to the expectation of alcohol than people diagnosed with the disorder, or healthy people without ...

Why we fail to understand our smartphone use

May 23, 2018
Checking your phone dozens of times a day indicates unconscious behaviour, which is "extremely repetitive" say psychologists.

Study confirms that men and women tend to adopt different navigation strategies

May 23, 2018
When navigating in a known environment, men prefer to take shortcuts to reach their destination more quickly, while women tend to use routes they know. This is according to Alexander Boone of UC Santa Barbara in the US who ...

Early life trauma in men associated with reduced levels of sperm microRNAs

May 22, 2018
Exposure to early life trauma can lead to poor physical and mental health in some individuals, which can be passed on to their children. Studies in mice show that at least some of the effects of stress can be transmitted ...

Training compassion 'muscle' may boost brain's resilience to others' suffering

May 22, 2018
It can be distressing to witness the pain of family, friends or even strangers going through a hard time. But what if, just like strengthening a muscle or learning a new hobby, we could train ourselves to be more compassionate ...

Study finds popular 'growth mindset' educational interventions aren't very effective

May 22, 2018
A new study co-authored by researchers at Michigan State University and Case Western Reserve University found that "growth mindset interventions," or programs that teach students they can improve their intelligence with effort—and ...

0 comments

Please sign in to add a comment. Registration is free, and takes less than a minute. Read more

Click here to reset your password.
Sign in to get notified via email when new comments are made.