Inherent envelope fluctuations in forward masking: Effects of age and hearing loss

Abstract

Forward masking is generally greater for Gaussian noise (GN) than for low-fluctuation noise maskers, i.e., GN disruption. Because the minimal hearing loss that is associated with older age may affect GN disruption differently than more significant hearing loss, the current study explored the contribution of minimal hearing loss associated with older age to GN disruption. GN disruption was measured using three masker-signal delays (25, 75, and 150 ms) for three adult groups: younger participants with normal hearing (NH), older participants with minimal hearing loss, and older participants with sensorineural hearing loss. The role of underlying mechanisms was tested using a computational model for midbrain neurons. The primary result suggests that older listeners with mild threshold elevations that typically occur with age may be more susceptible to the deleterious effects of masker envelope fluctuations than younger listeners with NH. Results from the computational model indicate that there may be a larger influence of efferent feedback and saturation of inner hair cells on forward masking and GN disruption than previously considered.

FIG. 1.

(Color online) Plots of hearing threshold for each participant group. Number of participants, age range, and M age are also provided. For this and the remaining box and whisker plot, each box represents the interquartile range, each line represents the median, each asterisk represents the M, each circle represents an outlier (>1.5 times the interquartile range), and whiskers represent the most extreme value that is not an outlier.

FIG. 2.

(Color online) Computational model, including MOC efferent system, showing example simulation of one frequency channel for OMHL. (a) GN and LFN stimuli at the input to the model ultimately result in (b) cochlear OHC gain factors that vary throughout the stimulus. (c) Low-spontaneous rate (LSR) and (d) high-spontaneous rate (HSR) AN responses are the instantaneous-rate functions at the output of the model synapse (right). (e) The model IC response of bandpass filter model is depicted. The LSR fiber was used to represent wide-dynamic-range responses from the cochlear nucleus, which were combined with the IC model response as inputs to the MOC stage (right). Larger fluctuations in the IC response to GN stimuli resulted in greater gain reduction throughout the GN masker as compared to the LFN masker. Therefore, the peak rate in response to the signal for the HSR fiber was lower after the GN masker than after the LFN masker. (f) Subsequently, for the IC same-frequency inhibition-excitation (SFIE) response, which was the input to the decision variable, the relative rates of the signal and masker were smaller for GN than for LFN. CF = 4 kHz. The signal level was 70 dB SPL.

FIG. 3.

Computational model of MOC efferent system, depicting mapping of input spike rate to MOC gain factor to cochlea. Input spike rate from IC was scaled and added to that from a LSR ANFs, low-pass filtered, and then converted to MOC gain factor. Higher MOC spike rate resulted in smaller gain factor.

FIG. 4.

(Color online) Plots of threshold with the LFN and GN maskers for each delay time. Absolute (Abs) thresholds are in the 150 ms panel. Threshold decreased as delay time increased.

FIG. 5.

(Color online) Plot of GN disruption. For the 25-ms masker-signal delay, GN disruption was larger for the OMHL participants than for the other two groups.

FIG. 6.

(Color online) Plot of OHC gain for each masker-signal delay. Responses are shown for one hair cell with CF = 4 kHz. Differences in OHC gain between GN (dashed lines) and LFN (solid lines) were greater for OMHL (circles/magenta) than for YNH (asterisks/blue) and OSNHL (squares/yellow) models. Symbols denote, in temporal sequence, start of the masker, end of the masker, and temporal center of the 10-ms signal. The signal level was set 10 dB above the predicted model threshold for the GN conditions (see Fig. 7).

FIG. 7.

(Color online) Plots of mean behavioral and model predicted thresholds for YNH, OMHL, and OSNHL with and without the efferent system. The masker-signal delay is indicated above each plot. For each participant group, LFN and GN thresholds are indicated for the left and right symbols, respectively.

FIG. 8.

(Color online) Plots of mean behavioral and model GN disruption for YNH, OMHL, and OSNHL with and without the efferent system. The masker-signal delay is indicated above each plot.