Speech analysis can help measure the diagnosis, severity, and timing of mental illness

Objective measurement of mental illness has long been challenging
.
However, according to a literature review in the January/February issue of the Harvard Psychiatric Review, there is ample evidence that analyzing language patterns can accurately diagnose depression and psychosis, measure their severity, and predict their onset
.
The magazine is published by Wolters Kluwer and is included in the Lippincott portfolio
.

This review reviewed the currently published literature on the use of speech pattern analysis to manage mental disorders and identified four key application areas: diagnostic classification, severity assessment, prognosis of morbidity, and prognosis and treatment outcomes
.
"A model that combines multiple speech features can very accurately distinguish between speakers with mental illness and healthy controls
," the researchers said.

Automated analysis is more promising than subjective measures such as interviews or questionnaires

Features of psychiatric disorders are often expressed through speech and language, and psychiatric clinical evaluation should consider patterns of speech, such as speech speed, coherence, and content
.
Advances in natural language processing, speech recognition, and computer science underscore the fact that it is possible
to use speech analysis as an objective, clinical measure of mental illness.

The research team reviewed hundreds of articles, papers, and reports on people with mental disorders that discussed various aspects
of their speech.
Case studies and studies of people with neurological disorders were excluded from the review
.
This included analyzing the passages
transcribed by the participants' speeches.

Most of the studies in this review that discuss the use of speech analysis in diagnosis have involved people with major depressive disorder, who tend to speak slowly, with pauses, negative content and lack of energy
.
In these studies, diagnostic accuracy was high, over 80%
in one study.

Automated analysis is also effective in predicting the onset of mental illness, especially in
high-risk populations.
Several studies have looked at speech semantics, including coherence and complexity, predicting the onset of psychosis two to two and a half years later with up to 100%
accuracy.
However, the literature on the effect of speech analysis on prognosis and treatment outcomes is limited and more research
is needed.

Importantly, the use of speech pattern analysis to assess suicide risk appears to have great potential
.
A recent study showed that by measuring variables such as unstable frequency, hesitation and nervousness, 73% of people were able to distinguish patients with suicidal ideation from healthy patients
.

Language differences, and other issues

There are a number of factors, such as drug effects, as well as demographic and cultural attributes (language, sex, and gender, etc.
) that can contribute to differences in speech patterns and make it challenging
to incorporate speaking into objective disease and outcome assessments.
In addition, the authors suggest that any further research should take into account changes in disease status over time, as most of the studies examined here were for patients who are currently ill, rather than whether similar patterns persist
long between symptoms.