Speech Perception Assessment and Training System

语音感知评估和训练系统

• 批准号: 6933696
• 负责人: James D Miller
• 金额: $ 40.37万
• 依托单位: COMMUNICATION DISORDERS TECHNOLOGY, INC
• 依托单位国家: 美国
• 财政年份: 2003
• 资助国家: 美国
• 起止时间: 2003-08-15 至 2007-04-30
• 项目状态: 已结题
• 来源: https://reporter.nih.gov/project-details/6933696
• 关键词: clinical research; clinical trial phase I; cochlear implants; computer assisted instruction; computer assisted patient care; computer system design /evaluation; deaf aid; deafness; educational resource design /development; hearing tests; human subject; human therapy evaluation; patient oriented research; phonology; speech recognition

DESCRIPTION (provided by applicant): The goal is a software Speech Perception Assessment and Training System (SPATS) for the aural rehabilitation of adult hearing aid and cochlear-implant users. The system will assess the user's ability to identify the constituents of English syllables; that is, syllable onsets, nuclei, and codas in quiet and in background noise, and it will provide individualized recognition training to improve these abilities. An administrative component of the system will keep and report records and simplify navigation of the system by clinicians and clients. Three parts of the project are: 1) the development of the assessment components of the system, 2) the development of the training components of the system, and 3) their clinical validation. Assessments will provide an inventory of a client's abilities to identify the constituents of English syllables. Including variation in talkers, phonetic contexts, and noise levels assures that the assessments are valid and objective evaluations of a listener's bottom-up, or analytic, speech-perception abilities. The training algorithm adaptively selects items and noise levels so that practice is concentrated on items of moderate difficulty weighted by their importance for speech understanding. Item importance is estimated from frequency of occurrence data. Further, learning is enhanced by providing many practice trials, immediate feedback, and, following errors, rapid comparisons of intended sounds with their confusors and by including both relatively easy, low-uncertainty and more difficult, high-uncertainty conditions. To better motivate the user and to encourage transfer to real-world speech, briefer periods of training on words and phrases will be regularly interspersed with the training on syllable constituents. Validation studies will measure how well the assessment system predicts success in identifying words and sentences, and general success in speech communication. Correlations with subjective estimates of benefit from, satisfaction with, and use of hearing aids and cochlear implants in relation to the proposed assessment tools will also be measured. Assessment components of the system will be suitable for patient evaluation and for the evaluation of the performance of hearing-aids and their fittings as well as for the evaluation of the performance of cochlear implants, maps, programs, and processing schemes. The effectiveness of the training system for improving the validation measures and, thus, the clients' overall ability to understand speech, will be measured. The training components of the system can be used in aural rehabilitation programs in clinical settings and in the home.

描述（由申请人提供）：目标是一个用于成人助听器和人工耳蜗用户听力康复的软件言语感知评估和训练系统（SPATS）。该系统将评估用户的能力，以确定英语音节的成分，即，音节起始，核，尾在安静和背景噪音，它将提供个性化的识别训练，以提高这些能力。该系统的一个管理组件将保存和报告记录，并简化临床医生和客户对系统的导航。该项目的三个部分是：1）系统评估组件的开发，2）系统培训组件的开发，以及3）其临床验证。 评估将提供一个清单的客户的能力，以确定英语音节的组成部分。包括谈话者，语音环境和噪音水平的变化，确保评估是有效的和客观的评价听者的自下而上，或分析，语音感知能力。训练算法自适应地选择项目和噪声水平，使实践集中在中等难度的项目加权的重要性，为语音理解。项目的重要性是从发生频率数据估计。此外，通过提供许多实践试验，即时反馈，以及在错误之后，将预期声音与其混淆器进行快速比较，以及通过包括相对容易的、低不确定性和更困难的、高不确定性的条件来增强学习。为了更好地激励用户并鼓励转移到真实世界的语音中，单词和短语的简短训练将定期与音节成分的训练穿插在一起。验证研究将衡量评估系统在识别单词和句子方面的预测成功率，以及语音交流的总体成功率。还将测量与拟议评估工具相关的助听器和人工耳蜗植入物的受益、满意度和使用的主观估计的相关性。该系统的评估组件将适用于患者评估和助听器及其配件的性能评估，以及人工耳蜗植入物、地图、程序和处理方案的性能评估。将衡量培训系统对改进验证措施的有效性，从而衡量客户理解语言的总体能力。该系统的训练组件可用于临床环境和家庭中的听觉康复计划。