Binaural unmasking of tones and speech in bilateral cochlear implantees

点击翻译按钮获取中文摘要

ABSTRACT: Binaural hearing can provide greatly improved signal detection and speech understanding in back- ground noise. However, bilateral cochlear-implant (BICI) users demonstrate limited binaural unmasking (i.e., improved detection and understanding) of signals using speech processors. The gap in knowledge is that it is unclear why binaural unmasking nearly disappears when using speech processors. Until the binaural benefit gap is better understood and reduced, BICI users will have a relatively poor quality of life because of difficulty local- izing sounds and communicating in background noise. The long-term goal is to understand and maximize bin- aural benefits imparted by BICIs. The objective in this application is to understand the breakdown of binaural unmasking as the stimulus complexity increases and when signals are presented via speech processors. The central hypothesis is that BICI users' binaural unmasking is primarily stunted by the interaural decorrelation caused by stimulus-encoding-driven factors (modulations, current spread, and compression). Most studies have shown underwhelming binaural unmasking in BICI users, but few have considered the limiting role of interaural decorrelation. Our approach to investigate this problem is two-pronged: to better understand and increase un- masking in BICI listeners, and better understand and intentionally decrease unmasking in NH listeners (i.e., making them more like the BICI listeners). Then, we will compare perceptual performance to stimulus analyses of the physical waveforms, the stimuli processed by speech processing algorithms, and modeled neural repre- sentations of the electrical signals. With strong preliminary data in hand, the central hypothesis will be tested by pursuing four specific aims: (1) Measure the extent to which the internal representation of modulated single- electrode stimuli in BICI listeners is interaurally decorrelated; (2) Measure the extent to which current spread and compression limit binaural unmasking of tones in noise in multi-electrode stimulation; (3) Develop a model of across-frequency static and dynamic ILD processing; and (4) Measure the extent to which binaural unmasking of speech in noise is reduced by interaural decorrelation from modulations, current spread, and compression. The approach is innovative because it (1) investigates signal-based factors affecting binaural unmasking and other binaural benefits in BICIs and (2) investigates the role of ILDs to impart these benefits. The proposed research is significant because it helps us understand the barriers to larger binaural benefits in BICI users, which can inform changes to speech processing strategies and clinical device setting guidelines. Larger binaural un- masking and benefits will produce an important positive impact on CI users' quality of life because they will be better able to localize sounds and communicate in noisy environments.