A Subcortical Model for Auditory Forward Masking with Efferent Control of Cochlear Gain

Abstract

Previous physiological and psychophysical studies have explored whether feedback to the cochlea from the efferent system influences forward masking. The present work proposes that the limited growth-of-masking (GOM) observed in auditory nerve (AN) fibers may have been misunderstood; namely, that this limitation may be due to the influence of anesthesia on the efferent system. Building on the premise that the unanesthetized AN may exhibit GOM similar to more central nuclei, the present computational modeling study demonstrates that feedback from the medial olivocochlear (MOC) efferents may contribute to GOM observed physiologically in onset-type neurons in both the cochlear nucleus and inferior colliculus (IC). Additionally, the computational model of MOC efferents used here generates a decrease in masking with longer masker-signal delays similar to that observed in IC physiology and in psychophysical studies. An advantage of this explanation over alternative physiological explanations (e.g., that forward masking requires inhibition from the superior paraolivary nucleus) is that this theory can explain forward masking observed in the brainstem, early in the ascending pathway. For explaining psychoacoustic results, one strength of this model is that it can account for the lack of elevation in thresholds observed when masker level is randomly varied from interval-to-interval, a result that is difficult to explain using the conventional temporal window model of psychophysical forward masking. Future directions for evaluating the efferent mechanism as a contributing mechanism for psychoacoustic results are discussed.

Keywords: auditory nerve; computational model; forward masking; inferior colliculus; medial olivocochlear reflex; temporal window.

Figure 1.

Schematic diagram of subcortical model with MOC efferents (adapted from Brennan et al., 2023). Cochlear gain is decreased if the input level (and therefore the WDR discharge rate) is increased. Cochlear gain is also decreased if IC discharge rate is increased by fluctuations in HSR fiber responses. The IC modulation filter model of Mao et al. (2013) projected to the MOC, while the final model used for calculating IC thresholds was the SFIE model (Nelson and Carney, 2004).

Figure 2.

Impact of tone masker level on response to probe tone. The level of a 4 kHz tone masker (left and right columns, 40 and 70 dB SPL) influences the response at various stages of the efferent model (blue). Responses of model without efferents are shown for comparison (gray). Rows show stimulus and five stages of the model: HSR AN fibers, WDR responses (modeled using LSR AN fibers), the gain factor or influence of both feedback pathways on cochlear gain over time; the response of the fluctuation pathway (mostly irrelevant for tone maskers that do not induce fluctuations); and the final response of the IC SFIE model. The WDR stage and feedback pathways are not included in the model without efferents. Note that, for the efferent model (blue), the HSR and IC responses to the probe tone (40 dB SPL, 10 ms, 2.8 ms delay) are substantially decreased for the higher-level masker in comparison with the probe tone responses for the lower-level masker condition, due to a change in cochlear gain. However, in the model without efferents (gray), the probe tone responses remain prominent for the 70 dB SPL masker.

Figure 3.

Decision-variable windows. IC SFIE model response to a 4 kHz, 40 dB SPL masker, 2.83 ms masker–probe delay, and 40 dB SPL, 10 ms probe tone (IC received input from model with efferents). Lines superimposed to show the placement of the physiological decision-variable window (bottom solid line) and psychophysical decision-variable window (top dotted line). The physiological decision-variable window was used for simulating the results of Ingham et al. (2016) and Nelson et al. (2009) and was shifted later or earlier depending on the masker–probe delay. The psychophysical decision-variable window was only used for simulating the paradigm of Jesteadt et al. (2005) and was not shifted with the masker–probe delay.

Figure 4.

Forward-masked thresholds versus masker level for models, physiology, and psychophysics. Growth of masking thresholds shown for IC and AN models (diamond and square markers, respectively), with and without efferents (blue and gray, respectively). IC physiology (black solid line) based on a linear fit to the average suppression in onset-type neurons from Ingham et al. (2016): 2 ms delay, 100 ms masker, 25 ms signal. Psychophysical GOM rate (slope of black dashed line) derived from Moore and Glasberg (1983): 0 ms delay, 210 ms masker, 20 ms signal. For physiology and models, masker level is plotted on the x-axis relative to the threshold of the masker in quiet. Probe suppression is plotted on the y-axis, in dB relative to the unmasked threshold of the probe. Psychophysical GOM curve is shifted along the x-axis to more closely match the SPL of model stimuli (SL and SPL were similar for the model because model unmasked thresholds were within 5 dB of 0 dB SPL). The IC physiological results shown here (from Ingham et al, 2016) were from experiments using urethane anesthesia, with potentially less impact on the efferent system than barbiturates (Guitton et al., 2004).

Figure 5.

Change in threshold with increasing masker/probe delay. Suppression at varying delays shown for IC and AN models (diamond and square markers, respectively), with and without efferents (blue and gray, respectively). IC physiology (black solid line): average suppression in awake recordings of IC neurons in Nelson et al. (2009); masker was 40 dB re neuron threshold (SL), 200 ms in duration, 20 ms signal. Psychophysics: from Moore and Glasberg (1983); masker at 40 dB SPL, 210 ms masker, 20 ms signal. Simulations matched parameters in Nelson et al. (2009); unmasked model thresholds were ∼1 dB SPL for the IC models and ∼5 dB SPL for the AN models, so 40 dB SPL and 40 dB SL were similar values for the models.

Figure 6.

Stable thresholds with masker-level rove. Psychophysical thresholds (black triangles; Jesteadt et al., 2005: experiment 1) and thresholds for IC model with efferent feedback (blue diamonds) shown for masker–probe delays of 0 and 30 ms, with masker level fixed at 70 dB SPL and with masker-level rove (filled and open symbols, respectively). Left panel, On-frequency condition (masker and probe at 4 kHz). Right panel, Off-frequency condition (masker at 2.4 kHz, probe at 4 kHz). Pairs of points with similar thresholds indicate the minimal impact of rove (x-coordinate offset to clearly show both filled and open symbols). Psychophysical suppression was calculated by subtracting a general estimate of unmasked thresholds for similar probe tones (15 dB SPL; Plack and Oxenham, 1998) from data reported in Jesteadt et al. (2005). Simulations matched parameters in Jesteadt et al. (2005).