INTELLIGIBILITY ASSESSMENT IN DYSARTHRIA

本研究的目标是开发改进的评估协议 为言语障碍提供定量分析评估 在患有神经系统疾病的儿童和成人中。 语言障碍 （构音障碍）将在肌萎缩侧索硬化症患者中进行研究。 硬化症，帕金森病，中风，小脑变性，脑 麻痹和发育性语言障碍。 改进的语音评估 可理解性是这项工作的一个特别重点，但语音问题 和语音质量。 要使用的方法是 标准临床评估（如评定量表）的组合， 可懂度评估和基于计算机的声学分析。 具体地，所述方法包括： 经验丰富的临床医生，可懂度的定量评估， 多参数声学分析，计算机语音校正 通过声信号的LPC再合成的异常，以及 从声学参数导出声道形状。 好各项工作 这些领域将基于来自大型 有构音障碍的人数。 这项研究的一个成果将 是包括可懂度分数的临床概况的库， 语音特征分析、语音/声音质量评级、声学 测量和神经学诊断。 将特别注意 受试者年龄和性别对特征的影响 构音障碍的神经学诊断。 评估协议 将在微型计算机上实现，并被设计成 临床实践 这项研究也将有助于发展 构音障碍分级和分类的专家系统。

The goal of this research is to develop improved assessment protocols that afford a quantitative and analytic evaluation of speech impairment in children and adults with neurological disorders. Speech impairments (dysarthrias) will be studied in persons with amyotrophic lateral sclerosis, Parkinson's disease, stroke, cerebellar degeneration, cerebral palsy, and developmental speech disorders. Improved evaluation of speech intelligibility is a particular focus of this work, but issues of speech and voice quality also are addressed. The methods to be used are a combination of standard clinical assessments (such as rating scales), intelligibility evaluations, and computer-based acoustic analyses. Specifically, the methods include: perceptual ratings of speech by experienced clinicians, quantitative assessment of intelligibility, a multiple-parameter acoustic analysis, computer correction of speech abnormalities through LPC resynthesis of the acoustic signal, and derivation of vocal tract shape from acoustic parameters. Work in all these areas will be based on recordings of speech samples from a large number of individuals with dysarthria. One product of the research will be a library of clinical profiles including intelligibility scores, phonetic feature analyses, ratings of speech/voice quality, acoustic measures, and neurological diagnosis. Particular attention will be given to the influences of subject age and sex on the characteristics of dysarthria for a given neurological diagnosis. The assessment protocols will be implemented on microcomputers and designed to be incorporated in clinical practice. The research also will contribute to the development of expert systems for the rating and classification of dysarthria.