Natural Language Processing markers in first episode psychosis and people at clinical high-risk

Recent work has suggested that disorganised speech might be a powerful predictor of later psychotic illness in clinical high risk subjects. To that end, several automated measures to quantify disorganisation of transcribed speech have been proposed. However, it remains unclear which measures are mos...

Full description

Saved in:
Bibliographic Details
Published in:Translational psychiatry 2021-12, Vol.11 (1), p.630-630, Article 630
Main Authors: Morgan, Sarah E., Diederen, Kelly, Vértes, Petra E., Ip, Samantha H. Y., Wang, Bo, Thompson, Bethany, Demjaha, Arsime, De Micheli, Andrea, Oliver, Dominic, Liakata, Maria, Fusar-Poli, Paolo, Spencer, Tom J., McGuire, Philip
Format: Article
Language:English
Subjects:
692/53/2421
692/699/476/1799
Automation
Behavioral Sciences
Biological Psychology
Biomarkers
Cognition
Humans
Medicine
Medicine & Public Health
Natural Language Processing
Neurosciences
Pharmacotherapy
Psychiatry
Psychosis
Psychotic Disorders - diagnosis
Speech
Citations: Items that this one cites
Items that cite this one
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:Recent work has suggested that disorganised speech might be a powerful predictor of later psychotic illness in clinical high risk subjects. To that end, several automated measures to quantify disorganisation of transcribed speech have been proposed. However, it remains unclear which measures are most strongly associated with psychosis, how different measures are related to each other and what the best strategies are to collect speech data from participants. Here, we assessed whether twelve automated Natural Language Processing markers could differentiate transcribed speech excerpts from subjects at clinical high risk for psychosis, first episode psychosis patients and healthy control subjects (total N  = 54). In-line with previous work, several measures showed significant differences between groups, including semantic coherence, speech graph connectivity and a measure of whether speech was on-topic, the latter of which outperformed the related measure of tangentiality. Most NLP measures examined were only weakly related to each other, suggesting they provide complementary information. Finally, we compared the ability of transcribed speech generated using different tasks to differentiate the groups. Speech generated from picture descriptions of the Thematic Apperception Test and a story re-telling task outperformed free speech, suggesting that choice of speech generation method may be an important consideration. Overall, quantitative speech markers represent a promising direction for future clinical applications.
ISSN:2158-3188
2158-3188
DOI:10.1038/s41398-021-01722-y