Progressive changes in descriptive discourse in First Episode Schizophrenia: a longitudinal computational semantics study

Abstract

Computational semantics, a branch of computational linguistics, involves automated meaning analysis that relies on how words occur together in natural language. This offers a promising tool to study schizophrenia. At present, we do not know if these word-level choices in speech are sensitive to the illness stage (i.e., acute untreated vs. stable established state), track cognitive deficits in major domains (e.g., cognitive control, processing speed) or relate to established dimensions of formal thought disorder. In this study, we collected samples of descriptive discourse in patients experiencing an untreated first episode of schizophrenia and healthy control subjects (246 samples of 1-minute speech; n = 82, FES = 46, HC = 36) and used a co-occurrence based vector embedding of words to quantify semantic similarity in speech. We obtained six-month follow-up data in a subsample (99 speech samples, n = 33, FES = 20, HC = 13). At baseline, semantic similarity was evidently higher in patients compared to healthy individuals, especially when social functioning was impaired; but this was not related to the severity of clinically ascertained thought disorder in patients. Across the study sample, higher semantic similarity at baseline was related to poorer Stroop performance and processing speed. Over time, while semantic similarity was stable in healthy subjects, it increased in patients, especially when they had an increasing burden of negative symptoms. Disruptions in word-level choices made by patients with schizophrenia during short 1-min descriptions are sensitive to interindividual differences in cognitive and social functioning at first presentation and persist over the early course of the illness.

Changes in the structure of spontaneous speech predict the disruption of hierarchical brain organization in first-episode psychosis

“…The graphs of both word-level and utterance-level meaning (Figure 4B) demonstrate a pattern of significantly higher centrality, weaker small-worldness, and a tendency for greater clustering. We capture this conceptually as a ‘shrinking’ or ‘narrowing’ semantic space, consistent with recent findings, which found a lesser average semantic distance (higher cosine semantic similarity) between consecutive words in FEP (19, 38). Higher centrality in these graphs suggests excessive connections between semantically distant units, while higher clustering indicates the formation of more semantic subdomains.…”

Widespread cortical thinning, excessive glutamate and impaired linguistic functioning in schizophrenia: A cluster analytic approach

“…Such NLP tools do not rely on a clinician's inferential skill to assess the cognitive-linguistic health status (Voleti et al, 2020) of patients from early stages of psychosis (Delvecchio et al, 2019) and are able to predict psychosis onset in individuals at clinical high-risk (CHR) (Bedi et al, 2015). These approaches have broadly focused on syntactic (Thomas et al, 1990;Thomas, 1996;Covington et al, 2005;Delvecchio et al, 2019) and semantic indices (Corcoran et al, 2018;Bar et al, 2019;Alonso-Sánchez et al, 2022;Parola et al, 2022) as the affected domains in psychosis.…”

Phenomenology as a resource for translational research in mental health: methodological trends, challenges and new directions

“…In addition to this initial proposal, other areas could be considered for phenomenological engagement on a larger scale. For example, natural language processing could be used to facilitate the analysis and management of large-scale phenomenological datasets (e.g., descriptive discourse in first-episode schizophrenia; Alonso-Sánchez et al ., 2022 ) and support early detection, prevention and treatment.…”