Central Factors in Auditory Masking

点击翻译按钮获取中文摘要

Project Summary / Abstract The goal of the work described in this proposal is to obtain a better understanding of the auditory communication difficulties experienced by listeners with sensorineural hearing loss. The proposed work examines "masking" - the interference in auditory communication caused by unwanted sounds - and how masking affects listeners with sensorineural hearing loss. The plan of research is based on the theory that masking consists of two main components: energetic masking, which is due primarily to interactions between target and masker sounds in the auditory periphery, and informational masking, which occurs at higher levels of processing despite an adequate representation of the target sounds in the periphery. The extent to which listeners with sensorineural hearing loss are affected by energetic masking versus informational masking in daily communication tasks - such as following a conversation in a crowded, noisy room - is not well known. The research plan examines conditions where speech masks other speech using methods that separate the influence of each type of masking. To accomplish this a series of psychophysical and speech recognition experiments are proposed in which energetic and informational masking are varied in a controlled manner. In the first aim, a quantitative estimate of performance after accounting for masking is planned using a signal processing approach - "ideal time-frequency segregation" - that reduces speech mixtures to target-dominated "glimpses". By comparing the performance of normal-hearing and hearing-impaired listeners in recognizing this processed stimulus, we will obtain a better understanding of whether hearing loss degrades performance because of an inability to synthesize the piecemeal glimpses of the target that occur in speech mixtures, or is due to increased energetic and/or informational masking. In the second aim, the role of linguistic factors in causing informational masking is examined and compared in normal-hearing and hearing-impaired listeners. Linguistic processing depends on an adequate representation of the speech signal in the periphery but also requires sufficient processing capacity for the tasks of segregating and attending to the speech stream under competition from other sounds including irrelevant speech. Explicit tests of how linguistic factors interact with lower-level segregation cues to cause - or release - informational masking will be conducted using a new question-and-answer paradigm that incorporates a measure of message comprehension. In the third aim, the ability of normal-hearing and hearing-impaired listeners to extract information from sequences of speech or learned nonspeech patterns will be compared under energetic and informational masking conditions. The hypothesis that hearing loss acts as an additional load on limited processing capacity will be examined in a series of novel masked recall experiments. Finally, throughout the experimental program measures of selected individual characteristics, abilities and prior experience will be obtained and correlated with the psychophysical and speech perception results in an attempt to learn the origins of the enormous differences often observed between individuals.