- Email: [email protected]
Efferent-induced change in human cochlear compression and its influence on masking of tones
Neuroscience Letters 485 (2010) 94–97
Contents lists available at ScienceDirect
Neuroscience Letters journal homepage: www.elsevier.com/locate/neulet
Efferent-induced change in human cochlear compression and its influence on masking of tones Shaum P. Bhagat ∗ , Paul H. Carter Hearing Science Laboratory, School of Audiology and Speech-Language Pathology, The University of Memphis, 807 Jefferson Avenue, Memphis, TN 38105, United States
a r t i c l e
i n f o
Article history: Received 15 July 2010 Accepted 25 August 2010 Keywords: Cochlea Compression Contralateral suppression Distortion-product otoacoustic emissions Masking Medial olivocochlear efferent neurons
a b s t r a c t Several lines of evidence suggest that medial olivocochlear (MOC) efferent neurons modify cochlear output to improve signal detection in noise. In animal models, stimulation of MOC efferents reduces the amount of compression in basilar membrane (BM) growth functions. Linearization of BM growth functions may assist in extending the neural response to the signal above that of noise, leading to a decrease in masking. In order to test this hypothesis, effects of MOC efferent neurons on BM compression were studied indirectly in humans by examining the effects of contralateral noise on distortion-product otoacoustic emission (DPOAE) input–output functions at 1.0 and 2.0 kHz. Compression threshold estimates from a three-segment linear regression model applied to the DPOAE functions were derived in order to determine correlations with psychophysical measurements of masking of tones at 1.0 and 2.0 kHz. Contralateral noise shifted the DPOAE compression threshold to a significantly higher level at 1.0 kHz, but not at 2.0 kHz. A significant negative correlation between the change in DPOAE compression threshold and the amount of masking at 1.0 kHz was observed, but no correlation between these variables was detected at 2.0 kHz. The results of this experiment at the lower test frequency indicated that contralateral noise linearized DPOAE input–output functions, and individuals with larger DPOAE compression threshold shifts tended to exhibit less masking. Under certain conditions, decreases in cochlear compression induced by MOC efferent neurons may lead to unmasking of tones presented in noise. © 2010 Elsevier Ireland Ltd. All rights reserved.
Outer hair cells (OHCs) in the mammalian cochlea receive innervation from crossed and uncrossed medial olivocochlear (MOC) efferent neurons originating near the superior olivary complex in the brainstem. The nonlinear transduction properties of OHCs have been implicated as the origin of the compressive input–output functions observed in the cochlea [17]. Although their functional role in auditory perception is not completely understood, several lines of evidence suggest that MOC efferent neurons can modify cochlear output to improve the detection of signals in noise. Stimulation of MOC efferent neurons: 1) hyperpolarizes OHCs and increases OHC conductance, possibly leading to a decrease in OHC motility [5,6], 2) reduces basilar membrane (BM) sensitivity and increases the slope of BM input–output functions for tones presented near the characteristic frequency (CF) of the recording location [14], 3) enhances neural output in the auditory nerve in response to brief CF tones at moderate and high levels presented in noise [7] and 4) increases the slope and dynamic range of neuronal input–output functions acquired in response to CF tones presented in noise for some neurons in the cochlear nucleus and inferior colliculus [13,15]. The decrease in cochlear sensitivity induced by MOC
∗ Corresponding author. Tel.: +1 901 678 5816; fax: +1 901 525 1282. E-mail address: [email protected] (S.P. Bhagat). 0304-3940/$ – see front matter © 2010 Elsevier Ireland Ltd. All rights reserved. doi:10.1016/j.neulet.2010.08.069
efferent neurons may aid detection of brief tones embedded in continuous noise by reducing neural adaptation brought on by the noise, resulting in an effective increase in the neural response to the tone [9]. Furthermore, the increased linearity of BM input–output functions at CF associated with activation of MOC efferent neurons may assist in extending the neural response to the tone above that of moderate to loud noise [14]. In humans, MOC efferent neurons have been studied indirectly by measuring distortion-product otoacoustic emissions (DPOAEs) in conditions with and without presentation of acoustic stimulation designed to activate MOC efferent neuronal activity. Believed to partly originate from nonlinear transduction properties of OHCs [16], DPOAEs are usually inaudible sounds that can be measured in the ear canal by a sensitive microphone. They are evoked by the simultaneous presentation of two tones (f1 and f2 ; f2 > f1 ) to an ear and the largest DPOAE in humans occurs at the 2f1 − f2 frequency [10]. During presentation of contralateral noise, DPOAE amplitudes can be suppressed, enhanced, or exhibit no changes relative to measurement conditions without the contralateral noise [1,3]. These changes seen in human DPOAE amplitudes are similar to MOC efferent-induced effects on the BM observed in animal models. Group DPOAE input–output functions in humans also exhibit linear and compressed segments that resemble the compressive BM input–output functions in animal models [4]. Measurement
S.P. Bhagat, P.H. Carter / Neuroscience Letters 485 (2010) 94–97
of DPOAEs over a range of stimulus levels therefore provides a means to indirectly assess the modification of cochlear compression by MOC efferent neurons in humans with the contralateral noise paradigm. A change in cochlear compression may then be related to some functional outcome, such as listener performance in detecting tones in noise. Some studies have shown small but significant associations between the amount of contralateral suppression of cochlear emissions evoked by tone pips and listener ability to detect tones in noise [11]. However, the influence of the MOC efferents on cochlear compression and masking of tones has not been completely clarified in humans. In this experiment, DPOAE input–output functions were fit with a three-segment linear regression model [19] in order to obtain compression estimates. The model contains linear-compressedlinear segments that are representative of BM input–output functions. The positions of the lower and upper breakpoints connecting the three segments and the slope of the compressed nonlinear segment are parameters in the model. These parameters were adjusted to provide the best fit (lowest rms error value) to the data. Based on the fitted functions, estimates of the lower breakpoints separating the low-level linear segment from the compressed middle segment were obtained. These lower breakpoints identify the stimulus level (in dB SPL) at which the function enters compression, and they are conceptually equivalent to compression thresholds. Conceivably, these DPOAE compression thresholds would register changes in compression induced by MOC efferent neurons activated by contralateral noise. The purpose of this experiment was 1) to examine if compression thresholds from DPOAE input–output functions acquired without contralateral noise were significantly different from compression thresholds from DPOAE input–output functions acquired with contralateral noise, and 2) to examine if there were significant relationships between the contralateral noise-induced change in DPOAE compression thresholds and the amount of simultaneous masking exhibited by individuals for selected listening conditions. Fourteen female adults with normal hearing (age 22–42 years) participated in the experiment. The procedures of this experiment were approved by the Institutional Review Board of the University of Memphis and written informed consent was obtained from all participants. The participants had hearing thresholds at or better than 20 dB HL in both ears for the standard audiometric test frequencies measured at inter-octave intervals from 0.25–8.0 kHz and type-A tympanograms bilaterally. Middle-ear muscle (MEM) reflex thresholds, obtained with a pulsed broadband noise activator presented to the ear contralateral to the test ear, were at or greater than 60 dB HL in both ears of every participant. Tone detection thresholds and DPOAE input–output functions were measured in each subject at 1.0 and 2.0 kHz. Contralateral noise effectively suppresses DPOAEs in this frequency range in humans, and these frequencies have been evaluated in previous work examining contralateral suppression and tone-in-noise detection [11,12]. All measurements occurred in the same double-walled, sound-treated enclosure. Thresholds were measured in odd-numbered subjects first, followed by DPOAE measurements. In even-numbered subjects, DPOAE measurements were completed first, followed by measurement of thresholds. The right ear of each participant was selected for measurement. All participants’ thresholds were measured with a Dell Optiflex GX 280 computer interfaced with Tucker-Davis Technologies (TDT) System 3 hardware and equipped with SykoFizX v. 2.0 software. Tonal stimuli were digitally generated (TDT, RP 2.1) at a nominal rate of 50 kHz. For measurement conditions with a masker, a broadband noise was generated at the same sampling rate as the tonal signals. Signals were sent to a TDT HB7 headphone buffer before being transduced by one of a pair of Sony MDR-V500 headphones. The duration of the tone signals and the noise masker was 300 ms, including 10 ms cosine-gated
95
Table 1 Absolute and masked thresholds (in dB SPL) at 1.0 and 2.0 kHz. The masker was a broadband noise presented at 60 dB SPL. Mean (standard deviation) values are shown. Thresholds
1.0 kHz
2.0 kHz
Absolute Masked
12.0 (3.4) 39.8 (1.9)
11.7 (4.8) 43.4 (1.5)
rise and fall times. The sequence of threshold testing was the same in each subject, with measurement of absolute thresholds in the quiet preceding measurement of masked thresholds in the presence of the noise masker. In masking conditions, the tone signal and the 60 dB noise masker were presented at the same time and were directed to the right ear. Estimation of threshold was provided by a two-interval, two-alternative forced-choice procedure with a two-down one-up adaptive rule that tracked the 71% correct performance level in each subject. Subjects voted by using a mouse to click on the selected interval icon on a computer monitor and feedback indicating the correct interval was provided. Each threshold run was comprised of 50 trials and threshold was defined as the average of the estimates obtained from three runs. Following the threshold procedure, the amount of masking (in dB) at each test frequency was calculated by subtracting the absolute threshold from the masked threshold in each participant. Table 1 lists mean absolute and masked thresholds at each test frequency. DPOAE input–output functions at f2 frequencies of 1.0 and 2.0 kHz were acquired in each subject using an Otodynamics ILO 296 analyzer interfaced with the same computer used in threshold testing. The f2 :f1 ratio was fixed at 1.22. Primary-tone levels at the higher frequency primary (L2 ) were incremented in 5-dB steps and were presented at targeted levels from 45 to 70 dB SPL. These levels were examined as MOC efferents effectively suppress and linearize BM vibrations within this range of stimulus levels [14]. Primary-tone levels at the lower frequency primary (L1 ) were calculated using the formula (L1 = 0.4L2 + 39) developed by Kummer et al. [8]. During data collection, the ILO probe was placed in the ear canal of the right ear and DPOAEs at 2f1 − f2 were measured for each primary-tone level. DPOAE measurements without contralateral noise were completed first. Next, an insert earphone was coupled to the contralateral (left) ear canal. Broadband noise bursts (duration = 600 ms, level = 60 dB SPL) generated by the TDT RP2.1 were presented while DPOAEs were measured in the right ear for each primary-tone level. Compression estimates were derived offline from DPOAE input–output functions fit with the threesegment linear regression model. Fitting of the model to the data was accomplished using the fminsearch function in MATLAB® . Fig. 1 illustrates DPOAE input–output functions and model fits to the data for two measurements obtained from one participant. The change in DPOAE compression threshold (in dB) between the no-noise and contralateral noise conditions was calculated for each participant separately for each test frequency. Kolmogorov–Smirnov tests evaluating the distribution of the dependent variables found that these variables were normally distributed. In order to determine if DPOAE compression thresholds from the no-noise condition were significantly different from DPOAE compression thresholds from the contralateral noise condition, paired-sample t-tests were separately conducted on compression threshold data at the 1.0 and 2.0 kHz test frequencies. Pearson product–moment correlations were calculated to examine the relationship between the change in DPOAE compression threshold and the amount of masking exhibited at each test frequency. Statistical analyses were performed using OriginPro 8.0 software. The level of significance selected for all tests was 0.05. The average rms error of the 56 model fits (14 subjects × 2 conditions × 2 frequencies) was 1.22 dB, indicating that overall,
96
S.P. Bhagat, P.H. Carter / Neuroscience Letters 485 (2010) 94–97
Fig. 1. DPOAE input–output functions. DPOAE data acquired from the ear of one participant is depicted. Open triangles represent DPOAE amplitudes in the condition without contralateral noise (left panel) and filled triangles represent DPOAE amplitudes in the condition with contralateral noise (right panel). The model fits (solid lines) and compression thresholds (open circles) are also shown.
the model performed well in fitting the DPOAE data. The mean compression thresholds identified by the model are shown in Fig. 2. Compression thresholds from DPOAE input–output functions acquired with contralateral noise were significantly higher (t = −2.57, d.f. = 13, p = 0.02) than compression thresholds from DPOAE input–output functions acquired without contralateral noise at 1.0 kHz. No significant difference (t = −0.81, d.f. = 13, p = 0.43) between the compression thresholds from the DPOAE input–output functions for the two conditions was detected at 2.0 kHz. Fig. 3 depicts the relationships between the change in DPOAE compression threshold (x-axis) and the amount of masking (y-axis) at each test frequency. The data at 1.0 kHz indicated the tendency for individuals exhibiting higher amounts of masking to have smaller changes in DPOAE compression thresholds compared to individuals exhibiting lower amounts of masking, who tended to have larger changes in DPOAE compression thresholds. In contrast, no clear trends in the data at 2.0 kHz were found. Statistical
Fig. 2. DPOAE compression thresholds. Mean DPOAE compression thresholds are shown at each test frequency. Open squares represent DPOAE compression thresholds acquired without contralateral noise and filled squares represent DPOAE compression thresholds acquired with contralateral noise. Error bars = ±1 standard error.
Fig. 3. Correlations. Amount of masking in dB (y-axis) is plotted against change in DPOAE compression threshold in dB (x-axis) at the 1.0 kHz (top) and 2.0 kHz (bottom) test frequencies. Lines of best fit determined by linear regression analyses are depicted.
testing confirmed these observations, as a significant correlation (r = −0.57, p = 0.03) between the change in DPOAE compression threshold and the amount of masking at 1.0 kHz was detected. The correlation between these two variables was not statistically significant (r = −0.05, p = 0.86) at 2.0 kHz. The effects of contralateral noise on DPOAEs in this experiment were qualitatively similar to the linearized BM input–output functions caused by stimulation of MOC efferent neurons in animal models. DPOAE compression thresholds were shifted to a higher level by the contralateral noise, indicating that the linear portion of DPOAE input–output functions at lower stimulus levels was extended presumably through the action of the MOC efferents. MEM reflexes elicited with noise activators were examined with clinical instrumentation in each participant in this study, and all subjects exhibited MEM reflex thresholds above the level of contralateral noise used in this experiment. Although sub-threshold MEM reflexes cannot be ruled out, Sun [18] concluded that for levels of contralateral noise below the MEM reflex, effects on DPOAEs are most likely mediated by the MOC reflex pathway. The threshold shifts induced by contralateral noise resulted in statistically significant differences between DPOAE compression thresholds at 1.0 kHz, but these differences were not statistically significant
S.P. Bhagat, P.H. Carter / Neuroscience Letters 485 (2010) 94–97
at 2.0 kHz. Previous work [12] has shown that contralateral suppression of DPOAEs with broadband noise is most effective at frequencies near 1.0 kHz and the amount of contralateral suppression declines at higher frequencies. The smaller contralateral noise-induced shifts in compression threshold at 2.0 kHz in this experiment may be related to less efficient MOC efferent function at the higher test frequency compared to the lower test frequency. However, a more recent study [2] found that in right ears, contralateral suppression of DPOAEs was greater at 2.0 kHz than at 1.0 kHz, although the differences in contralateral suppression between the frequencies were on the order of about 1 dB. Another factor to consider is that in this experiment, noise bursts were used as contralateral suppressors instead of continuous noise. Noise bursts were selected because they are more similar in duration to the electric pulse trains used to stimulate MOC efferent neurons in animal models [14,15] than continuous noise. Different results, including larger shifts in compression thresholds at both test frequencies, may have been observed if continuous noise was used as the contralateral suppressor. There was a statistically significant negative correlation between the change in DPOAE compression threshold and the amount of masking at 1.0 kHz. Individuals with large shifts in DPOAE compression thresholds to higher levels, possibly due to more efficient MOC efferent function, exhibited less masking than individuals with smaller compression threshold shifts. These findings were consistent with the speculation that MOC efferent neurons decrease BM compression to extend the neural response to the signal above that of moderate noise [14]. The fact that no correlation between variables at 2.0 kHz occurred may due to the fact that in most participants, little or no change in DPOAE compression thresholds were observed and this made it difficult to detect a significant relationship between the variables. In conclusion, the findings of this experiment suggested that contralateral noise can linearize DPOAE input–output functions, resulting in DPOAE compression thresholds being shifted to higher levels. At the lower test frequency, individuals with larger changes in DPOAE compression thresholds exhibited less masking. A decrease in cochlear compression induced by MOC efferent neurons may contribute to unmasking of tones presented in noise. However, these results should be interpreted cautiously, given the small sample size in this experiment. Efforts to characterize the effects of the MOC efferents on cochlear compression will undoubtedly lead to better understanding of how listeners perceive signals in noise. Acknowledgments The authors thank Dr. Enrique Lopez-Poveda for his assistance with the three-segment linear regression model.
97
References [1] C. Abdala, S.K. Mishra, T.L. Williams, Considering distortion product otoacoustic emission fine structure in measurements of the medial olivocochlear reflex, J. Acoust. Soc. Am. 125 (2009) 1584–1594. [2] S.R. Atcherson, M.J. Martin, R. Lintvedt, Contralateral noise has possible asymmetric frequency-sensitive effect on the 2f1 − f2 otoacoustic emission in humans, Neurosci. Lett. 438 (2008) 107–110. [3] S.P. Bhagat, C.A. Champlin, Evaluation of distortion products produced by the human auditory system, Hear. Res. 193 (2004) 51–67. [4] P. Boege, T. Janssen, Pure-tone threshold estimation from extrapolated distortion product otoacoustic emission I/O-functions in normal and cochlear hearing loss ears, J. Acoust. Soc. Am. 111 (2002) 1810–1818. [5] M.G. Evans, Acetylcholine activates two currents in guinea-pig outer hair cells, J. Physiol. 491 (1996) 563–578. [6] G.D. Housley, J.F. Ashmore, Direct measurement of the action of acetylcholine on isolated outer hair cells of the guinea pig cochlea, Proc. Biol. Sci. 244 (1991) 161–167. [7] T. Kawase, B. Delgutte, M.C. Liberman, Antimasking effects of the olivocochlear reflex. II. Enhancement of auditory-nerve response to masked tones, J. Neurophysiol. 70 (1993) 2533–2549. [8] P. Kummer, T. Janssen, W. Arnold, The level and growth behavior of the 2f1 − f2 distortion product otoacoustic emission and its relationship to auditory sensitivity in normal hearing and cochlear hearing loss, J. Acoust. Soc. Am. 103 (1998) 3431–3444. [9] M.C. Liberman, J.J. Guinan, Feedback control of the auditory periphery: antimasking effects of middle-ear muscles vs. olivocochlear efferents, J. Commun. Disord. 31 (1998) 471–482. [10] B. Lonsbury-Martin, G.K. Martin, Otoacoustic emissions, in: A.F. Jahn, J. SantosSacchi (Eds.), Physiology of the Ear, Singular, San Diego, 2001, pp. 443– 480. [11] C. Micheyl, T. Morlet, A.L. Giraud, L. Collet, A. Morgon, Contralateral suppression of evoked otoacoustic emissions and detection of a multi-tone complex in noise, Acta Otolaryngol. 115 (1995) 178–182. [12] A. Moulin, L. Collet, R. Duclaux, Contralateral auditory stimulation alters acoustic distortion products in humans, Hear. Res. 68 (1993) 193–210. [13] W.H. Mulders, K. Seluakumaran, D. Robertson, Effects of centrifugal pathways on responses of cochlear nucleus neurons to signals in noise, Eur. J. Neurosci. 27 (2008) 702–714. [14] I.J. Russell, E. Murugasu, Medial efferent inhibition suppresses basilar membrane responses to near characteristic frequency tones of moderate to high intensities, J. Acoust. Soc. Am. 102 (1997) 1734–1738. [15] K. Seluakumaran, W.H. Mulders, D. Robertson, Unmasking effects of olivocochlear efferent activation on responses of inferior colliculus neurons, Hear. Res. 243 (2008) 35–46. [16] C.A. Shera, Mechanisms of mammalian otoacoustic emission and their implication for the clinical utility of otoacoustic emissions, Ear. Hear. 25 (2004) 86–97. [17] A. Stasiunas, A. Verikas, P. Kemesis, M. Bacauskiene, N. Stasiuniene, K. Malmqvist, A non-linear circuit for simulating OHC of the cochlea, Med. Eng. Phys. 25 (2003) 591–601. [18] X.M. Sun, Contralateral suppression of distortion product otoacoustic emissions and the middle-ear muscle reflex in human ears, Hear. Res. 237 (2008) 66– 75. [19] I. Yasin, C.J. Plack, The effects of a high-frequency suppressor on tuning curves and derived basilar-membrane response functions, J. Acoust. Soc. Am. 114 (2003) 322–332.
Contents lists available at ScienceDirect
Neuroscience Letters journal homepage: www.elsevier.com/locate/neulet
Efferent-induced change in human cochlear compression and its influence on masking of tones Shaum P. Bhagat ∗ , Paul H. Carter Hearing Science Laboratory, School of Audiology and Speech-Language Pathology, The University of Memphis, 807 Jefferson Avenue, Memphis, TN 38105, United States
a r t i c l e
i n f o
Article history: Received 15 July 2010 Accepted 25 August 2010 Keywords: Cochlea Compression Contralateral suppression Distortion-product otoacoustic emissions Masking Medial olivocochlear efferent neurons
a b s t r a c t Several lines of evidence suggest that medial olivocochlear (MOC) efferent neurons modify cochlear output to improve signal detection in noise. In animal models, stimulation of MOC efferents reduces the amount of compression in basilar membrane (BM) growth functions. Linearization of BM growth functions may assist in extending the neural response to the signal above that of noise, leading to a decrease in masking. In order to test this hypothesis, effects of MOC efferent neurons on BM compression were studied indirectly in humans by examining the effects of contralateral noise on distortion-product otoacoustic emission (DPOAE) input–output functions at 1.0 and 2.0 kHz. Compression threshold estimates from a three-segment linear regression model applied to the DPOAE functions were derived in order to determine correlations with psychophysical measurements of masking of tones at 1.0 and 2.0 kHz. Contralateral noise shifted the DPOAE compression threshold to a significantly higher level at 1.0 kHz, but not at 2.0 kHz. A significant negative correlation between the change in DPOAE compression threshold and the amount of masking at 1.0 kHz was observed, but no correlation between these variables was detected at 2.0 kHz. The results of this experiment at the lower test frequency indicated that contralateral noise linearized DPOAE input–output functions, and individuals with larger DPOAE compression threshold shifts tended to exhibit less masking. Under certain conditions, decreases in cochlear compression induced by MOC efferent neurons may lead to unmasking of tones presented in noise. © 2010 Elsevier Ireland Ltd. All rights reserved.
Outer hair cells (OHCs) in the mammalian cochlea receive innervation from crossed and uncrossed medial olivocochlear (MOC) efferent neurons originating near the superior olivary complex in the brainstem. The nonlinear transduction properties of OHCs have been implicated as the origin of the compressive input–output functions observed in the cochlea [17]. Although their functional role in auditory perception is not completely understood, several lines of evidence suggest that MOC efferent neurons can modify cochlear output to improve the detection of signals in noise. Stimulation of MOC efferent neurons: 1) hyperpolarizes OHCs and increases OHC conductance, possibly leading to a decrease in OHC motility [5,6], 2) reduces basilar membrane (BM) sensitivity and increases the slope of BM input–output functions for tones presented near the characteristic frequency (CF) of the recording location [14], 3) enhances neural output in the auditory nerve in response to brief CF tones at moderate and high levels presented in noise [7] and 4) increases the slope and dynamic range of neuronal input–output functions acquired in response to CF tones presented in noise for some neurons in the cochlear nucleus and inferior colliculus [13,15]. The decrease in cochlear sensitivity induced by MOC
∗ Corresponding author. Tel.: +1 901 678 5816; fax: +1 901 525 1282. E-mail address: [email protected] (S.P. Bhagat). 0304-3940/$ – see front matter © 2010 Elsevier Ireland Ltd. All rights reserved. doi:10.1016/j.neulet.2010.08.069
efferent neurons may aid detection of brief tones embedded in continuous noise by reducing neural adaptation brought on by the noise, resulting in an effective increase in the neural response to the tone [9]. Furthermore, the increased linearity of BM input–output functions at CF associated with activation of MOC efferent neurons may assist in extending the neural response to the tone above that of moderate to loud noise [14]. In humans, MOC efferent neurons have been studied indirectly by measuring distortion-product otoacoustic emissions (DPOAEs) in conditions with and without presentation of acoustic stimulation designed to activate MOC efferent neuronal activity. Believed to partly originate from nonlinear transduction properties of OHCs [16], DPOAEs are usually inaudible sounds that can be measured in the ear canal by a sensitive microphone. They are evoked by the simultaneous presentation of two tones (f1 and f2 ; f2 > f1 ) to an ear and the largest DPOAE in humans occurs at the 2f1 − f2 frequency [10]. During presentation of contralateral noise, DPOAE amplitudes can be suppressed, enhanced, or exhibit no changes relative to measurement conditions without the contralateral noise [1,3]. These changes seen in human DPOAE amplitudes are similar to MOC efferent-induced effects on the BM observed in animal models. Group DPOAE input–output functions in humans also exhibit linear and compressed segments that resemble the compressive BM input–output functions in animal models [4]. Measurement
S.P. Bhagat, P.H. Carter / Neuroscience Letters 485 (2010) 94–97
of DPOAEs over a range of stimulus levels therefore provides a means to indirectly assess the modification of cochlear compression by MOC efferent neurons in humans with the contralateral noise paradigm. A change in cochlear compression may then be related to some functional outcome, such as listener performance in detecting tones in noise. Some studies have shown small but significant associations between the amount of contralateral suppression of cochlear emissions evoked by tone pips and listener ability to detect tones in noise [11]. However, the influence of the MOC efferents on cochlear compression and masking of tones has not been completely clarified in humans. In this experiment, DPOAE input–output functions were fit with a three-segment linear regression model [19] in order to obtain compression estimates. The model contains linear-compressedlinear segments that are representative of BM input–output functions. The positions of the lower and upper breakpoints connecting the three segments and the slope of the compressed nonlinear segment are parameters in the model. These parameters were adjusted to provide the best fit (lowest rms error value) to the data. Based on the fitted functions, estimates of the lower breakpoints separating the low-level linear segment from the compressed middle segment were obtained. These lower breakpoints identify the stimulus level (in dB SPL) at which the function enters compression, and they are conceptually equivalent to compression thresholds. Conceivably, these DPOAE compression thresholds would register changes in compression induced by MOC efferent neurons activated by contralateral noise. The purpose of this experiment was 1) to examine if compression thresholds from DPOAE input–output functions acquired without contralateral noise were significantly different from compression thresholds from DPOAE input–output functions acquired with contralateral noise, and 2) to examine if there were significant relationships between the contralateral noise-induced change in DPOAE compression thresholds and the amount of simultaneous masking exhibited by individuals for selected listening conditions. Fourteen female adults with normal hearing (age 22–42 years) participated in the experiment. The procedures of this experiment were approved by the Institutional Review Board of the University of Memphis and written informed consent was obtained from all participants. The participants had hearing thresholds at or better than 20 dB HL in both ears for the standard audiometric test frequencies measured at inter-octave intervals from 0.25–8.0 kHz and type-A tympanograms bilaterally. Middle-ear muscle (MEM) reflex thresholds, obtained with a pulsed broadband noise activator presented to the ear contralateral to the test ear, were at or greater than 60 dB HL in both ears of every participant. Tone detection thresholds and DPOAE input–output functions were measured in each subject at 1.0 and 2.0 kHz. Contralateral noise effectively suppresses DPOAEs in this frequency range in humans, and these frequencies have been evaluated in previous work examining contralateral suppression and tone-in-noise detection [11,12]. All measurements occurred in the same double-walled, sound-treated enclosure. Thresholds were measured in odd-numbered subjects first, followed by DPOAE measurements. In even-numbered subjects, DPOAE measurements were completed first, followed by measurement of thresholds. The right ear of each participant was selected for measurement. All participants’ thresholds were measured with a Dell Optiflex GX 280 computer interfaced with Tucker-Davis Technologies (TDT) System 3 hardware and equipped with SykoFizX v. 2.0 software. Tonal stimuli were digitally generated (TDT, RP 2.1) at a nominal rate of 50 kHz. For measurement conditions with a masker, a broadband noise was generated at the same sampling rate as the tonal signals. Signals were sent to a TDT HB7 headphone buffer before being transduced by one of a pair of Sony MDR-V500 headphones. The duration of the tone signals and the noise masker was 300 ms, including 10 ms cosine-gated
95
Table 1 Absolute and masked thresholds (in dB SPL) at 1.0 and 2.0 kHz. The masker was a broadband noise presented at 60 dB SPL. Mean (standard deviation) values are shown. Thresholds
1.0 kHz
2.0 kHz
Absolute Masked
12.0 (3.4) 39.8 (1.9)
11.7 (4.8) 43.4 (1.5)
rise and fall times. The sequence of threshold testing was the same in each subject, with measurement of absolute thresholds in the quiet preceding measurement of masked thresholds in the presence of the noise masker. In masking conditions, the tone signal and the 60 dB noise masker were presented at the same time and were directed to the right ear. Estimation of threshold was provided by a two-interval, two-alternative forced-choice procedure with a two-down one-up adaptive rule that tracked the 71% correct performance level in each subject. Subjects voted by using a mouse to click on the selected interval icon on a computer monitor and feedback indicating the correct interval was provided. Each threshold run was comprised of 50 trials and threshold was defined as the average of the estimates obtained from three runs. Following the threshold procedure, the amount of masking (in dB) at each test frequency was calculated by subtracting the absolute threshold from the masked threshold in each participant. Table 1 lists mean absolute and masked thresholds at each test frequency. DPOAE input–output functions at f2 frequencies of 1.0 and 2.0 kHz were acquired in each subject using an Otodynamics ILO 296 analyzer interfaced with the same computer used in threshold testing. The f2 :f1 ratio was fixed at 1.22. Primary-tone levels at the higher frequency primary (L2 ) were incremented in 5-dB steps and were presented at targeted levels from 45 to 70 dB SPL. These levels were examined as MOC efferents effectively suppress and linearize BM vibrations within this range of stimulus levels [14]. Primary-tone levels at the lower frequency primary (L1 ) were calculated using the formula (L1 = 0.4L2 + 39) developed by Kummer et al. [8]. During data collection, the ILO probe was placed in the ear canal of the right ear and DPOAEs at 2f1 − f2 were measured for each primary-tone level. DPOAE measurements without contralateral noise were completed first. Next, an insert earphone was coupled to the contralateral (left) ear canal. Broadband noise bursts (duration = 600 ms, level = 60 dB SPL) generated by the TDT RP2.1 were presented while DPOAEs were measured in the right ear for each primary-tone level. Compression estimates were derived offline from DPOAE input–output functions fit with the threesegment linear regression model. Fitting of the model to the data was accomplished using the fminsearch function in MATLAB® . Fig. 1 illustrates DPOAE input–output functions and model fits to the data for two measurements obtained from one participant. The change in DPOAE compression threshold (in dB) between the no-noise and contralateral noise conditions was calculated for each participant separately for each test frequency. Kolmogorov–Smirnov tests evaluating the distribution of the dependent variables found that these variables were normally distributed. In order to determine if DPOAE compression thresholds from the no-noise condition were significantly different from DPOAE compression thresholds from the contralateral noise condition, paired-sample t-tests were separately conducted on compression threshold data at the 1.0 and 2.0 kHz test frequencies. Pearson product–moment correlations were calculated to examine the relationship between the change in DPOAE compression threshold and the amount of masking exhibited at each test frequency. Statistical analyses were performed using OriginPro 8.0 software. The level of significance selected for all tests was 0.05. The average rms error of the 56 model fits (14 subjects × 2 conditions × 2 frequencies) was 1.22 dB, indicating that overall,
96
S.P. Bhagat, P.H. Carter / Neuroscience Letters 485 (2010) 94–97
Fig. 1. DPOAE input–output functions. DPOAE data acquired from the ear of one participant is depicted. Open triangles represent DPOAE amplitudes in the condition without contralateral noise (left panel) and filled triangles represent DPOAE amplitudes in the condition with contralateral noise (right panel). The model fits (solid lines) and compression thresholds (open circles) are also shown.
the model performed well in fitting the DPOAE data. The mean compression thresholds identified by the model are shown in Fig. 2. Compression thresholds from DPOAE input–output functions acquired with contralateral noise were significantly higher (t = −2.57, d.f. = 13, p = 0.02) than compression thresholds from DPOAE input–output functions acquired without contralateral noise at 1.0 kHz. No significant difference (t = −0.81, d.f. = 13, p = 0.43) between the compression thresholds from the DPOAE input–output functions for the two conditions was detected at 2.0 kHz. Fig. 3 depicts the relationships between the change in DPOAE compression threshold (x-axis) and the amount of masking (y-axis) at each test frequency. The data at 1.0 kHz indicated the tendency for individuals exhibiting higher amounts of masking to have smaller changes in DPOAE compression thresholds compared to individuals exhibiting lower amounts of masking, who tended to have larger changes in DPOAE compression thresholds. In contrast, no clear trends in the data at 2.0 kHz were found. Statistical
Fig. 2. DPOAE compression thresholds. Mean DPOAE compression thresholds are shown at each test frequency. Open squares represent DPOAE compression thresholds acquired without contralateral noise and filled squares represent DPOAE compression thresholds acquired with contralateral noise. Error bars = ±1 standard error.
Fig. 3. Correlations. Amount of masking in dB (y-axis) is plotted against change in DPOAE compression threshold in dB (x-axis) at the 1.0 kHz (top) and 2.0 kHz (bottom) test frequencies. Lines of best fit determined by linear regression analyses are depicted.
testing confirmed these observations, as a significant correlation (r = −0.57, p = 0.03) between the change in DPOAE compression threshold and the amount of masking at 1.0 kHz was detected. The correlation between these two variables was not statistically significant (r = −0.05, p = 0.86) at 2.0 kHz. The effects of contralateral noise on DPOAEs in this experiment were qualitatively similar to the linearized BM input–output functions caused by stimulation of MOC efferent neurons in animal models. DPOAE compression thresholds were shifted to a higher level by the contralateral noise, indicating that the linear portion of DPOAE input–output functions at lower stimulus levels was extended presumably through the action of the MOC efferents. MEM reflexes elicited with noise activators were examined with clinical instrumentation in each participant in this study, and all subjects exhibited MEM reflex thresholds above the level of contralateral noise used in this experiment. Although sub-threshold MEM reflexes cannot be ruled out, Sun [18] concluded that for levels of contralateral noise below the MEM reflex, effects on DPOAEs are most likely mediated by the MOC reflex pathway. The threshold shifts induced by contralateral noise resulted in statistically significant differences between DPOAE compression thresholds at 1.0 kHz, but these differences were not statistically significant
S.P. Bhagat, P.H. Carter / Neuroscience Letters 485 (2010) 94–97
at 2.0 kHz. Previous work [12] has shown that contralateral suppression of DPOAEs with broadband noise is most effective at frequencies near 1.0 kHz and the amount of contralateral suppression declines at higher frequencies. The smaller contralateral noise-induced shifts in compression threshold at 2.0 kHz in this experiment may be related to less efficient MOC efferent function at the higher test frequency compared to the lower test frequency. However, a more recent study [2] found that in right ears, contralateral suppression of DPOAEs was greater at 2.0 kHz than at 1.0 kHz, although the differences in contralateral suppression between the frequencies were on the order of about 1 dB. Another factor to consider is that in this experiment, noise bursts were used as contralateral suppressors instead of continuous noise. Noise bursts were selected because they are more similar in duration to the electric pulse trains used to stimulate MOC efferent neurons in animal models [14,15] than continuous noise. Different results, including larger shifts in compression thresholds at both test frequencies, may have been observed if continuous noise was used as the contralateral suppressor. There was a statistically significant negative correlation between the change in DPOAE compression threshold and the amount of masking at 1.0 kHz. Individuals with large shifts in DPOAE compression thresholds to higher levels, possibly due to more efficient MOC efferent function, exhibited less masking than individuals with smaller compression threshold shifts. These findings were consistent with the speculation that MOC efferent neurons decrease BM compression to extend the neural response to the signal above that of moderate noise [14]. The fact that no correlation between variables at 2.0 kHz occurred may due to the fact that in most participants, little or no change in DPOAE compression thresholds were observed and this made it difficult to detect a significant relationship between the variables. In conclusion, the findings of this experiment suggested that contralateral noise can linearize DPOAE input–output functions, resulting in DPOAE compression thresholds being shifted to higher levels. At the lower test frequency, individuals with larger changes in DPOAE compression thresholds exhibited less masking. A decrease in cochlear compression induced by MOC efferent neurons may contribute to unmasking of tones presented in noise. However, these results should be interpreted cautiously, given the small sample size in this experiment. Efforts to characterize the effects of the MOC efferents on cochlear compression will undoubtedly lead to better understanding of how listeners perceive signals in noise. Acknowledgments The authors thank Dr. Enrique Lopez-Poveda for his assistance with the three-segment linear regression model.
97
References [1] C. Abdala, S.K. Mishra, T.L. Williams, Considering distortion product otoacoustic emission fine structure in measurements of the medial olivocochlear reflex, J. Acoust. Soc. Am. 125 (2009) 1584–1594. [2] S.R. Atcherson, M.J. Martin, R. Lintvedt, Contralateral noise has possible asymmetric frequency-sensitive effect on the 2f1 − f2 otoacoustic emission in humans, Neurosci. Lett. 438 (2008) 107–110. [3] S.P. Bhagat, C.A. Champlin, Evaluation of distortion products produced by the human auditory system, Hear. Res. 193 (2004) 51–67. [4] P. Boege, T. Janssen, Pure-tone threshold estimation from extrapolated distortion product otoacoustic emission I/O-functions in normal and cochlear hearing loss ears, J. Acoust. Soc. Am. 111 (2002) 1810–1818. [5] M.G. Evans, Acetylcholine activates two currents in guinea-pig outer hair cells, J. Physiol. 491 (1996) 563–578. [6] G.D. Housley, J.F. Ashmore, Direct measurement of the action of acetylcholine on isolated outer hair cells of the guinea pig cochlea, Proc. Biol. Sci. 244 (1991) 161–167. [7] T. Kawase, B. Delgutte, M.C. Liberman, Antimasking effects of the olivocochlear reflex. II. Enhancement of auditory-nerve response to masked tones, J. Neurophysiol. 70 (1993) 2533–2549. [8] P. Kummer, T. Janssen, W. Arnold, The level and growth behavior of the 2f1 − f2 distortion product otoacoustic emission and its relationship to auditory sensitivity in normal hearing and cochlear hearing loss, J. Acoust. Soc. Am. 103 (1998) 3431–3444. [9] M.C. Liberman, J.J. Guinan, Feedback control of the auditory periphery: antimasking effects of middle-ear muscles vs. olivocochlear efferents, J. Commun. Disord. 31 (1998) 471–482. [10] B. Lonsbury-Martin, G.K. Martin, Otoacoustic emissions, in: A.F. Jahn, J. SantosSacchi (Eds.), Physiology of the Ear, Singular, San Diego, 2001, pp. 443– 480. [11] C. Micheyl, T. Morlet, A.L. Giraud, L. Collet, A. Morgon, Contralateral suppression of evoked otoacoustic emissions and detection of a multi-tone complex in noise, Acta Otolaryngol. 115 (1995) 178–182. [12] A. Moulin, L. Collet, R. Duclaux, Contralateral auditory stimulation alters acoustic distortion products in humans, Hear. Res. 68 (1993) 193–210. [13] W.H. Mulders, K. Seluakumaran, D. Robertson, Effects of centrifugal pathways on responses of cochlear nucleus neurons to signals in noise, Eur. J. Neurosci. 27 (2008) 702–714. [14] I.J. Russell, E. Murugasu, Medial efferent inhibition suppresses basilar membrane responses to near characteristic frequency tones of moderate to high intensities, J. Acoust. Soc. Am. 102 (1997) 1734–1738. [15] K. Seluakumaran, W.H. Mulders, D. Robertson, Unmasking effects of olivocochlear efferent activation on responses of inferior colliculus neurons, Hear. Res. 243 (2008) 35–46. [16] C.A. Shera, Mechanisms of mammalian otoacoustic emission and their implication for the clinical utility of otoacoustic emissions, Ear. Hear. 25 (2004) 86–97. [17] A. Stasiunas, A. Verikas, P. Kemesis, M. Bacauskiene, N. Stasiuniene, K. Malmqvist, A non-linear circuit for simulating OHC of the cochlea, Med. Eng. Phys. 25 (2003) 591–601. [18] X.M. Sun, Contralateral suppression of distortion product otoacoustic emissions and the middle-ear muscle reflex in human ears, Hear. Res. 237 (2008) 66– 75. [19] I. Yasin, C.J. Plack, The effects of a high-frequency suppressor on tuning curves and derived basilar-membrane response functions, J. Acoust. Soc. Am. 114 (2003) 322–332.