Audibility as a Predictor of Speech Recognition and Listening Effort

描述(由申请人提供)：拟议研究的目标是基于语音可理解性指数(SII)改进对儿童语音识别和听力努力的预测。SII是一个单一的数字，用于量化不同听力条件下语音刺激的平均可听度，通常用于优化儿童助听器设置。SII计算基于频率-重要性权重(FIW)，该权重对特定频率范围对语音识别的贡献进行数值计算。尽管之前的研究表明，儿童比成年人需要更多的高频信息来实现最佳的语音理解，但基于听力正常的成年人的数据的FIW目前被用于估计儿童的语音识别。研究表明，基于成人的预测高估了儿童的表现，这可能会导致临床医生选择不能为言语和语言发展提供最佳基础的助听器特征。在以前的研究中，没有对儿童进行专门的FIW测量，儿童的听觉和听力努力之间的关系仍然没有确定。在拟议研究的目标1中，将对听力正常的成年人和儿童进行FIW测量，以努力改进基于SII的儿童语音识别预测。在目标2中，来自儿童的FIW将被用来预测一组新的听力正常儿童的语音识别和听力努力。假设成人和儿童的FIW将有所不同，儿童FIW将导致比当前SII更准确地预测语音识别和听力努力。 公共卫生相关性：当前提案的结果将加强对听力损失儿童的助听器信号处理的选择，以专注于最大限度地理解言语和最大限度地减少听力努力。我们对成人和儿童使用光谱线索进行言语感知的能力的不同之处的理解也将受益于当前提案的结果。进一步改进听力损失儿童的放大策略可能会在言语和语言发展方面产生更好的结果。

DESCRIPTION (provided by applicant): The goal of the proposed research is to improve predictions of speech recognition and listening effort for children based on the Speech Intelligibility Index (SII). The SII is a single number used to quantify the average audibility of speech stimuli under different listening conditions and is often used to optimize hearing aid settings for children. SII calculations are based on frequency-importance weights (FIW), which place a numerical value on the contribution of a specific frequency range to speech recognition. Despite previous studies demonstrating that children require more high-frequency information than adults to achieve optimal speech understanding, FIW based on data from normal- hearing adults are currently used to estimate speech recognition of children. Studies have suggested that adult-based predictions overestimate performance for children, which could lead clinicians to select hearing-aid characteristics that do not provide an optimal basis for speech and language development. FIW have not been specifically measured with children in previous studies, and the relationship between audibility and listening effort in children remains undetermined. In Aim 1 of the proposed research, FIW will be measured for normal-hearing adults and children in an effort to improve SII- based predictions of speech recognition for children. In Aim 2, the child-derived FIW will be used to predict speech recognition and listening effort in a new group of normal-hearing children. It is hypothesized that FIW will differ for adults and children and that child-derived FIW will result in more accurate predictions of speech recognition and listening effort than the current SII. PUBLIC HEALTH RELEVANCE: The results of the current proposal will enhance selection of hearing-aid signal processing for children with hearing loss to focus on maximizing speech understanding and minimizing listening effort. Our understanding of how adults and children differ in their ability to use spectral cues for speech perception will also benefit from the results of the current proposal. Further improvements in amplification strategies for children with hearing loss could result in better outcomes in speech and language development.