Natural language processing methods are sensitive to sub-clinical linguistic differences in schizophrenia spectrum disorders

Abstract Computerized natural language processing (NLP) allows for objective and sensitive detection of speech disturbance, a hallmark of schizophrenia spectrum disorders (SSD). We explored several methods for characterizing speech changes in SSD (n = 20) compared to healthy control (HC) participant...

Descripción completa

Guardado en:
Detalles Bibliográficos
Autores principales: Sunny X. Tang, Reno Kriz, Sunghye Cho, Suh Jung Park, Jenna Harowitz, Raquel E. Gur, Mahendra T. Bhati, Daniel H. Wolf, João Sedoc, Mark Y. Liberman
Formato: article
Lenguaje:EN
Publicado: Nature Portfolio 2021
Materias:
Acceso en línea:https://doaj.org/article/5719c8246cbb4adb929bfe059cb11ecc
Etiquetas: Agregar Etiqueta
Sin Etiquetas, Sea el primero en etiquetar este registro!
id oai:doaj.org-article:5719c8246cbb4adb929bfe059cb11ecc
record_format dspace
spelling oai:doaj.org-article:5719c8246cbb4adb929bfe059cb11ecc2021-12-02T15:43:16ZNatural language processing methods are sensitive to sub-clinical linguistic differences in schizophrenia spectrum disorders10.1038/s41537-021-00154-32334-265Xhttps://doaj.org/article/5719c8246cbb4adb929bfe059cb11ecc2021-05-01T00:00:00Zhttps://doi.org/10.1038/s41537-021-00154-3https://doaj.org/toc/2334-265XAbstract Computerized natural language processing (NLP) allows for objective and sensitive detection of speech disturbance, a hallmark of schizophrenia spectrum disorders (SSD). We explored several methods for characterizing speech changes in SSD (n = 20) compared to healthy control (HC) participants (n = 11) and approached linguistic phenotyping on three levels: individual words, parts-of-speech (POS), and sentence-level coherence. NLP features were compared with a clinical gold standard, the Scale for the Assessment of Thought, Language and Communication (TLC). We utilized Bidirectional Encoder Representations from Transformers (BERT), a state-of-the-art embedding algorithm incorporating bidirectional context. Through the POS approach, we found that SSD used more pronouns but fewer adverbs, adjectives, and determiners (e.g., “the,” “a,”). Analysis of individual word usage was notable for more frequent use of first-person singular pronouns among individuals with SSD and first-person plural pronouns among HC. There was a striking increase in incomplete words among SSD. Sentence-level analysis using BERT reflected increased tangentiality among SSD with greater sentence embedding distances. The SSD sample had low speech disturbance on average and there was no difference in group means for TLC scores. However, NLP measures of language disturbance appear to be sensitive to these subclinical differences and showed greater ability to discriminate between HC and SSD than a model based on clinical ratings alone. These intriguing exploratory results from a small sample prompt further inquiry into NLP methods for characterizing language disturbance in SSD and suggest that NLP measures may yield clinically relevant and informative biomarkers.Sunny X. TangReno KrizSunghye ChoSuh Jung ParkJenna HarowitzRaquel E. GurMahendra T. BhatiDaniel H. WolfJoão SedocMark Y. LibermanNature PortfolioarticlePsychiatryRC435-571ENnpj Schizophrenia, Vol 7, Iss 1, Pp 1-8 (2021)
institution DOAJ
collection DOAJ
language EN
topic Psychiatry
RC435-571
spellingShingle Psychiatry
RC435-571
Sunny X. Tang
Reno Kriz
Sunghye Cho
Suh Jung Park
Jenna Harowitz
Raquel E. Gur
Mahendra T. Bhati
Daniel H. Wolf
João Sedoc
Mark Y. Liberman
Natural language processing methods are sensitive to sub-clinical linguistic differences in schizophrenia spectrum disorders
description Abstract Computerized natural language processing (NLP) allows for objective and sensitive detection of speech disturbance, a hallmark of schizophrenia spectrum disorders (SSD). We explored several methods for characterizing speech changes in SSD (n = 20) compared to healthy control (HC) participants (n = 11) and approached linguistic phenotyping on three levels: individual words, parts-of-speech (POS), and sentence-level coherence. NLP features were compared with a clinical gold standard, the Scale for the Assessment of Thought, Language and Communication (TLC). We utilized Bidirectional Encoder Representations from Transformers (BERT), a state-of-the-art embedding algorithm incorporating bidirectional context. Through the POS approach, we found that SSD used more pronouns but fewer adverbs, adjectives, and determiners (e.g., “the,” “a,”). Analysis of individual word usage was notable for more frequent use of first-person singular pronouns among individuals with SSD and first-person plural pronouns among HC. There was a striking increase in incomplete words among SSD. Sentence-level analysis using BERT reflected increased tangentiality among SSD with greater sentence embedding distances. The SSD sample had low speech disturbance on average and there was no difference in group means for TLC scores. However, NLP measures of language disturbance appear to be sensitive to these subclinical differences and showed greater ability to discriminate between HC and SSD than a model based on clinical ratings alone. These intriguing exploratory results from a small sample prompt further inquiry into NLP methods for characterizing language disturbance in SSD and suggest that NLP measures may yield clinically relevant and informative biomarkers.
format article
author Sunny X. Tang
Reno Kriz
Sunghye Cho
Suh Jung Park
Jenna Harowitz
Raquel E. Gur
Mahendra T. Bhati
Daniel H. Wolf
João Sedoc
Mark Y. Liberman
author_facet Sunny X. Tang
Reno Kriz
Sunghye Cho
Suh Jung Park
Jenna Harowitz
Raquel E. Gur
Mahendra T. Bhati
Daniel H. Wolf
João Sedoc
Mark Y. Liberman
author_sort Sunny X. Tang
title Natural language processing methods are sensitive to sub-clinical linguistic differences in schizophrenia spectrum disorders
title_short Natural language processing methods are sensitive to sub-clinical linguistic differences in schizophrenia spectrum disorders
title_full Natural language processing methods are sensitive to sub-clinical linguistic differences in schizophrenia spectrum disorders
title_fullStr Natural language processing methods are sensitive to sub-clinical linguistic differences in schizophrenia spectrum disorders
title_full_unstemmed Natural language processing methods are sensitive to sub-clinical linguistic differences in schizophrenia spectrum disorders
title_sort natural language processing methods are sensitive to sub-clinical linguistic differences in schizophrenia spectrum disorders
publisher Nature Portfolio
publishDate 2021
url https://doaj.org/article/5719c8246cbb4adb929bfe059cb11ecc
work_keys_str_mv AT sunnyxtang naturallanguageprocessingmethodsaresensitivetosubclinicallinguisticdifferencesinschizophreniaspectrumdisorders
AT renokriz naturallanguageprocessingmethodsaresensitivetosubclinicallinguisticdifferencesinschizophreniaspectrumdisorders
AT sunghyecho naturallanguageprocessingmethodsaresensitivetosubclinicallinguisticdifferencesinschizophreniaspectrumdisorders
AT suhjungpark naturallanguageprocessingmethodsaresensitivetosubclinicallinguisticdifferencesinschizophreniaspectrumdisorders
AT jennaharowitz naturallanguageprocessingmethodsaresensitivetosubclinicallinguisticdifferencesinschizophreniaspectrumdisorders
AT raquelegur naturallanguageprocessingmethodsaresensitivetosubclinicallinguisticdifferencesinschizophreniaspectrumdisorders
AT mahendratbhati naturallanguageprocessingmethodsaresensitivetosubclinicallinguisticdifferencesinschizophreniaspectrumdisorders
AT danielhwolf naturallanguageprocessingmethodsaresensitivetosubclinicallinguisticdifferencesinschizophreniaspectrumdisorders
AT joaosedoc naturallanguageprocessingmethodsaresensitivetosubclinicallinguisticdifferencesinschizophreniaspectrumdisorders
AT markyliberman naturallanguageprocessingmethodsaresensitivetosubclinicallinguisticdifferencesinschizophreniaspectrumdisorders
_version_ 1718385807562113024