Speech in psychosis has long been ascribed as involving 'loosening of associations'. We pursued the aim to elucidate its underlying cognitive mechanisms by analysing picture descriptions from 94 subjects (29 healthy controls, 18 participants at clinical high risk, 29 with first-episode psychosis, and 18 with chronic schizophrenia), using five language models with different computational architectures: FastText, which represents meaning non-contextually/statically; BERT, which represents contextual meaning sensitive to grammar and context; Infersent and SBERT, which provide sentential representations; and CLIP, which evaluates speech relative to a visual stimulus. These models were used to quantify semantic distances crossed between successive tokens/sentences, and semantic perplexity indicating unexpectedness in continuations. Results showed that, among patients, semantic similarity increased when measured with FastText, Infersent, and SBERT, while it decreased with CLIP and BERT. Higher perplexity was observed in first-episode psychosis. Static semantic measures were associated with clinically measured impoverishment of thought and referential semantic measures with disorganization. These patterns indicate a shrinking conceptual semantic space as represented by static language models, which co-occurs with a widening in the referential semantic space as represented by contextual models. This duality underlines the need to separate these two forms of meaning for understanding mechanisms involved in semantic change in psychosis.
Large language models provide high-dimensional representations (embeddings) of word meaning, which allow quantifying changes in the geometry of the semantic space in mental disorders. A pattern of a more condensed (‘shrinking’) semantic space marked by an increase in mean semantic similarity between words has been recently documented in psychosis across several languages. We aimed to explore this pattern further in picture descriptions provided by a transdiagnostic German sample of patients with schizophrenia spectrum disorders (SSD) (n = 42), major depression (MDD, n = 43), and healthy controls (n = 44). Compared to controls, both clinical groups showed more restricted dynamic navigational patterns as captured by the time series of semantic distances crossed, while also showing differential patterns in the total distances and trajectories navigated. These findings demonstrate alterations centred on the dynamics of the flow of meaning across the semantic space in SSD and MDD, preserving previous indications towards a shrinking semantic space in both cases.
11 September 2023 in American Journal of Speech-Language Pathology
Background: Decline in language has emerged as a new potential biomarker for the early detection of Alzheimer's disease (AD). It remains unclear how sensitive language measures are across different tasks, language domains, and languages, and to what extent changes can be reliably detected in early stages such as subjective cognitive decline (SCD) and mild cognitive impairment (MCI). Method: Using a scene construction task for speech elicitation in a new Spanish/Catalan speaking cohort (N = 119), we automatically extracted features across seven domains, three acoustic (spectral, cepstral, and voice quality), one prosodic, and three from text (morpholexical, semantic, and syntactic). They were forwarded to a random forest classifier to evaluate the discriminability of participants with probable AD dementia, amnestic and nonamnestic MCI, SCD, and cognitively healthy controls. Repeated-measures analyses of variance and paired-samples Wilcoxon signed-ranks test were used to assess whether and how performance differs significantly across groups and linguistic domains. Results: The performance scores of the machine learning classifier were generally satisfactorily high, with the highest scores over.9. Model performance was significantly different for linguistic domains (p < .001), and speech versus text (p = .043), with speech features outperforming textual features, and voice quality performing best. High diagnostic classification accuracies were seen even within both cognitively healthy (controls vs. SCD) and MCI (amnestic and nonamnestic) groups. Conclusion: Speech-based machine learning is powerful in detecting cognitive decline and probable AD dementia across a range of different feature domains, though important differences exist between these domains as well.
Psychosis implicates changes across a broad range of cognitive functions. These functions are cortically organized in the form of a hierarchy ranging from primary sensorimotor (unimodal) to higher‐order association cortices, which involve functions such as language (transmodal). Language has long been documented as undergoing structural changes in psychosis. We hypothesized that these changes as revealed in spontaneous speech patterns may act as readouts of alterations in the configuration of this unimodal‐to‐transmodal axis of cortical organization in psychosis. Results from 29 patients with first‐episodic psychosis (FEP) and 29 controls scanned with 7 T resting‐state fMRI confirmed a compression of the cortical hierarchy in FEP, which affected metrics of the hierarchical distance between the sensorimotor and default mode networks, and of the hierarchical organization within the semantic network. These organizational changes were predicted by graphs representing semantic and syntactic associations between meaningful units in speech produced during picture descriptions. These findings unite psychosis, language, and the cortical hierarchy in a single conceptual scheme, which helps to situate language within the neurocognition of psychosis and opens the clinical prospect for mental dysfunction to become computationally measurable in spontaneous speech.
Background and Hypothesis Identifying schizophrenia spectrum disorders (SSD) from spontaneous speech features is a key focus in computational psychiatry today. Study Design We present a task-voting procedure using different speech-elicitation tasks to predict SSD in Spanish, followed by ablation studies highlighting the roles of specific tasks and feature domains. Speech from five tasks was recorded from 92 subjects (49 with SSD and 41 controls). A total of 319 features were automatically extracted, from which 24 were pre-selected based on between-feature correlations and ANOVA F-values, covering acoustic-prosody, morphosyntax, and semantic similarity metrics. Study Results ExtraTrees-based classification using these features yielded an accuracy of 0.840 on hold-out data. Ablating picture descriptions impaired performance most, followed by story reading, retelling, and free speech. Removing morphosyntactic measures impaired performance most, followed by acoustic and semantic measures. Mixed-effect models suggested significant group differences on all 24 features. In SSD, speech patterns were slower and more variable temporally, while variations in pitch, amplitude, and sound intensity decreased. Semantic similarity between speech and prompts decreased, while minimal distances from embedding centroids to each word increased, and word-to-word similarity arrays became more predictable, all replicating patterns documented in other languages. Morphosyntactically, SSD patients used more first-person pronouns together with less third-person pronouns, and more punctuations and negations. Semantic metrics correlated with a range of positive symptoms, and multiple acoustic-prosodic features with negative symptoms. Conclusions This study highlights the importance of combining different speech tasks and features for SSD detection, and validates previously found patterns in psychosis for Spanish.