Language as a biomarker in those at high-risk for psychosis.
Language as a biomarker in those at high-risk for psychosis.
复制标题
语言作为精神病高危人群的生物标志物。
DOI：
10.1016/j.schres.2015.04.023
复制
发表时间：
2015
影响因子：
4.5
通讯作者：
Elvevåg,B
中科院分区：
文献类型：
--
作者：
Rosenstein,M;Foltz,PW;DeLisi,LE;Elvevåg,B
这项研究讨论了随后发展为严重精神疾病以及延误治疗会导致不良预后的问题。我们通过得出一组衡量连贯、语义和句法规则的语言特征来分析每个参与者的回答中使用的语言。使用Akaike信息标准的逐步评估为每个TAT卡建立预测模型，以选择语言度量来预测组成员。检查五个TAT预测模型中的每一个中的语言特征，产生了类似于前面描述的分组。我们将这些发现作为概念的进一步证明，即结合语音的句法结构和语义内容足以产生极其微妙的语言偏离，这可能为寻找可以为早期干预提供信息的精神病生物标志物提供一个框架。回到我们的假设，自动化技术可以检测人力资源参与者和对照组之间的差异。然而，这是否归因于所选择的提示或语音和书面文本形式之间的差异，需要调查。(数据库记录(C)2017 APA，保留所有权利)
Presents the study which discuses the subsequently develop severe mental illness and that delaying treatment results in poor prognosis. We analyzed the language used in each participant's response by deriving a set of language features that measured coherence, semantics and syntactic regularities. A predictive model was built for each TAT card using stepwise evaluation of the Akaike information criterion to select language measures to predict group membership. Examining the language features in each of the five TAT predictive models, yielded groupings similar to those previously described. We present these findings as a further proof of concept that combining syntactic structure and semantic content of speech is sufficiently rich to yield extremely subtle language deviances that may provide a framework for searching for biomarkers of psychosis that can inform early intervention. Returning to our hypotheses, automated techniques can detect differences between HR participants and controls. However, whether this is attributable to the prompts chosen or differences between speech versus written text modalities requires investigation.(PsycINFO Database Record (c) 2017 APA, all rights reserved)