Page 35 - Volume 12, Issue 2 - Spring 2012

PHYSIOLOGICAL CORRELATES OF PERCEPTUAL DEFICITS FOLLOWING SENSORINEURAL HEARING LOSS
Michael G. Heinz
Speech, Language, and Hearing Sciences & Biomedical Engineering Purdue University
West Lafayette, Indiana 47907
Introduction
“The translational implications of perceptual deficits in temporal processing with sensorineural hearing loss depend critically on their physiological basis, which unfortunately remains unknown.”
Several translational issues related
to sensorineural hearing loss
(SNHL) have recently re-invigor-
ated an active debate in the field of psy-
chological and physiological acoustics
regarding the perceptual role of the
temporal structure of sound. Much of
this recent interest over the last decade
has centered on the mathematical fact
that any sound can be separated into
slowly-varying temporal envelope
(ENV) and rapidly varying temporal
fine-structure (TFS) components (Fig.
1). This dichotomy has been motivated
in part from cochlear implants, which
are neural prostheses that can restore
the sense of hearing for patients with
through electrical stimulation of remaining cochlear neu- rons. Although often quite successful in understanding speech in quiet conditions, cochlear-implant patients often struggle to understand speech in noisy conditions or to appreciate music. These limitations have motivated numer- ous perceptual studies to understand the relative impor- tance of ENV and TFS cues because cochlear implants cur- rently only provide slowly varying ENV cues and are unable to provide rapidly varying TFS cues.
Envelope cues have been shown to be sufficient for understanding speech, even in a minimal number of fre- quency bands (Shannon et al., 1995; Smith et al., 2002). Fine structure has been argued to be useful for music per- ception, sound localization, lexical-tone perception, and source segregation (Smith et al., 2002; Xu and Pfingst, 2003; Qin and Oxenham, 2005). Such findings have been given as motivation for efforts to develop cochlear-implant strate- gies to provide fine-structure in addition to envelope cues (e.g., Rubinstein et al., 1999; Nie et al., 2005). It is also believed that the use of temporal fine structure relates to the remarkable ability of normal-hearing listeners to “listen in the dips” of background sounds, e.g., in the presence of con- current talkers (Moore and Glasberg, 1987; Schooneveldt and Moore, 1987).
In contrast, listeners with SNHL often face great diffi- culty in understanding speech in the presence of temporal- ly varying backgrounds (Peters et al., 1998), and this diffi- culty is only partially resolved with advanced amplification algorithms (Moore et al., 1999). A number of recent per- ceptual studies suggest that listeners with SNHL have a reduced ability to use TFS cues, and that this deficit may be correlated with their inability to listen in the dips (Lorenzi
et al., 2006; Moore et al., 2006; Hopkins and Moore, 2007). These results have fueled an active debate about the role of temporal coding in normal and impaired hearing, and have important implications for improving the ability of hearing aids and cochlear implants to restore speech perception in noise. However, the translational implications of per- ceptual deficits in temporal processing with SNHL depend critically on their physiological basis, which unfortu- nately remains unknown. The most straightforward hypothesis is that auditory-nerve fibers, which provide all of the information at the output of
the cochlea (Fig. 2), are no longer able to encode the rapid- ly varying TFS information following SNHL. Prior to our recent studies, the physiological evidence regarding audi- tory nerve (AN) fiber phase locking with SNHL was con- flicting (Harrison and Evans, 1979; Woolf et al., 1981; Miller et al., 1997). Thus, a number of important questions exist: (1) are TFS coding deficits peripheral or central, and (2) is it the quantity and/or quality of TFS coding that is diminished following SNHL?
This article describes recent work in our lab exploring the physiological bases for these perceptual results through a combination of neurophysiological, perceptual, and com- putational modeling approaches. Specifically, we have been exploring three hypotheses for the physiological bases of perceptual TFS deficits: (1) the temporal coding within individual AN fibers is degraded in quantity and/or quality following SNHL, (2) even if there is not diminished phase locking in individual fibers, the relative temporal coding across AN fibers may be degraded in ways that diminish spatio-temporal cues that have been proposed to be percep- tually relevant, and (3) the perceptual TFS deficits may not actually reflect diminished TFS coding in the periphery due to “artifacts” that complicate the separation of TFS and ENV cues within the cochlea. Overall, our results suggest that the fundamental ability of fibers to phase lock to tem- poral fine structure is not degraded by SNHL; however, a number of other effects of SNHL have been observed that may contribute to perceptual deficits in temporal process- ing of complex stimuli. These results have important trans- lational implications for auditory prostheses because they suggest that perceptual “fine-structure deficits” do not sim- ply reflect degraded phase locking, but likely involve more complicated coding mechanisms.
34 Acoustics Today, April 2012
profound deafness