Speech and Language analytics for early detection of psychosis 30007428

Project: Research project

Project Details

Description

Speech and Language analytics for early detection of psychosis 30007428 Speech and Language analytics for early detection of psychosis Speech and Language analytics for early detection of psychosis Structural, semantic, and acoustic analysis of speech in patients with a high risk of psychosis Introduction Since the initial descriptions of psychosis and schizophrenia, speech/language impairments are in the core symptomatology that describes the phenomenology associated with this mental suffering (some grouped in a set of symptoms described as formal thought disorder -FTD - accessed through speech/language). In recent years, different natural language processing tools (NLP) have efficiently measured aspects of FTD, helping the prediction of negative and cognitive symptomatology linked to psychosis, and the early identification of schizophrenia (even during the first episode), or conversion to psychosis (at the prodromal phase). There is major potential in learning which aspects of the FTD phenomenology, including the characteristic cognitive impairment, are associated with those NLP measurements in different clinical contexts, such as during the prodromal phase. Hypotheses According to the literature in schizophrenia, we hypothesize that negative symptomatology, as well as cognitive performance will be associated with NLP markers, and predictive of remission of the attenuated symptoms. In addition, we expect that the rate and rhythm of speech will be associated with negative and positive symptoms. Furthermore, acoustic features correlated with pause in speech will be related to negative features such as avolition; whereas the increased rate of speech will be correlated with incoherence. Objectives The aim of this study is to understand the phenomenological and cognitive associations of NLP metrics (structural, semantic and acoustic levels) from a short oral narrative protocol in the context of clinical high risk for psychosis - CHR, and to verify the validity of these speech/language markers for prediction of remission of attenuated psychotic symptomatology.
StatusActive
Effective start/end date12/15/1612/31/22

Funding

  • INDUSTRY: Foreign Company: $272,797.00

Fingerprint

Explore the research topics touched on by this project. These labels are generated based on the underlying awards/grants. Together they form a unique fingerprint.