An automated method to analyze language use in patients with schizophrenia and their first-degree relatives

Abstract

Communication disturbances are prevalent in schizophrenia, and since it is a heritable illness these are likely present -albeit in a muted form -in the relatives of patients. Given the time-consuming, and often subjective nature of discourse analysis, these deviances are frequently not assayed in large scale studies. Recent work in computational linguistics and statistical-based semantic analysis has shown the potential and power of automated analysis of communication. We present an automated and objective approach to modeling discourse that detects very subtle deviations between probands, their first-degree relatives and unrelated healthy controls. Although these findings should be regarded as preliminary due to the limitations of the data at our disposal, we present a brief analysis of the models that best differentiate these groups in order to illustrate the utility of the method for future explorations of how language components are differentially affected by familial and illness related issues.

Prediction of psychosis across protocols and risk cohorts using automated language analysis

“…In our prior small natural language processing study 15 , we found two measures of syntactic complexity -shorter sentences and reduced use of determiner pronouns that introduce dependent clauses -to be both predictive of psychosis and highly correlated with negative symptoms. In the present study, the failure of sentence length to predict psychosis in the training dataset may be a consequence of the brief and structured responses that were elicited (<20 mean words per response) 12 , as compared with prior studies (>120 mean words/response 15 , 800 words/response 12 and >10 sentences/response 30 ).…”

“…Hoffman applied manual discourse analysis to transcribed speech from schizophrenia patients, finding a reduction in semantic coherence 28 , a finding replicated later using computer-assisted discourse analysis 29 . It has only been in the last decade that natural language processing linguistic corpus-based analyses, specifically latent semantic analysis, have been applied to language production in schizophrenia, finding decreases in semantic coherence that correlate with clinical ratings, functional impairment, and task-related activation in language circuits 11,12 . Now, in the two CHR studies to date, latent semantic analysis with machine learning has shown decreased semantic coherence to predict subsequent psychosis onset.…”

“…More commonly found in schizophrenia speech is a reduction in syntactic complexity 27,32 , typically operationalized as shorter sentences, and most evident when open-ended narrative is elicited 12,30,31,33 . In our prior small natural language processing study 15 , we found two measures of syntactic complexity -shorter sentences and reduced use of determiner pronouns that introduce dependent clauses -to be both predictive of psychosis and highly correlated with negative symptoms.…”

“…Grouping of successive word-vectors can be used to estimate the semantic coherence of a narrative. Latent semantic analysis has been applied to speech in schizophrenia, finding an association of decreased semantic coherence with clinical ratings of thought disorder and functional impairment, and with abnormal task-related activation in language circuits 11,12 . …”

Detecting order–disorder transitions in discourse: Implications for schizophrenia

“…Indeed, such studies -using Latent Semantic Analysis (LSA) which models and matches discourse content (Landauer and Dumais, 1997;Landauer et al, 2007) -have demonstrated that it is possible to evaluate patients with schizophrenia based on open-ended verbalizations. These automatically derived language scores have distinguished patients from controls accurately (and patients from other patients, and also from their family members), using both large discourse samples as well as responses consisting of only a few words (Elvevåg et al, 2007(Elvevåg et al, , 2010.…”