Memoire Online - Contrainte Psycho-Physiques et Electrophysiologiques sur le codage de la stimulation électrique chez les sujets porteurs d'un implant cochléaire

Cet article compare la mesure du seuil différentiel en intensité (IDL) aux amplitudes et latences des PEAP pour une population normo-entendante (N=8) et une population implantée cochléaire (N=6).

Chez les sujets normo-entendants, les IDL ainsi que l'amplitude des ondes II et III varient de manière non-monotone en fonction de l'intensité de stimulation. Ces modifications non-linéaires peuvent s'expliquer par des seuils de déclenchement différents des fibres afférentes de type I au niveau de la cochlée (Liberman, 1978)

Chez les sujets implantés cochléaires, les IDL ainsi que l'amplitude des ondes II et III varient de manière monotone en fonction de l'intensité de stimulation.

L'absence de mécanismes complexes de la biomécanique cochléaire et du recrutement des fibres à différents seuils d'intensité lors d'une stimulation électrique via l'implant cochléaire Digisonic rend la fonction de IDL facilement calculable en fonction de l'amplitude des ondes.

Intensity Discrimination and Auditory Brainstem Responses
in Cochlear Implant and Normal-Hearing Listeners

Intensity-discrimination limens (IDLs) and auditory brainstem responses (ABRs) were measured as a function of stimulus intensity in 6 cochlear implant (CI) and 8 normal-hearing (NH) listeners. Pulse-train stimuli were delivered electrically to the auditory nerve in CI listeners and acoustically in NH listeners. In CI listeners, the IDLs expressed as Weber fractions decreased monotonically with increasing intensity. In NH listeners, a nonmonotonic IDL function showing a peak a midintensities was observed. ABR wave amplitudes increased regularly with intensity only in CI listeners. Results support the notion that the slight decrease in Weber's fractions with increasing sound intensity--generally referred to as "the near-miss to Weber's law"--is subtended by retrocochlear processes, whereas the increase in Weber's fractions at midlevels--known as "the severe departure from Weber's law"--originates in cochlear mechanisms.

The encoding of intensity in the auditory system has been a matter of intense debate over the past decades, and various neurophysiological mechanisms have been put forward to account for the characteristics of auditory intensity perception evidenced by behavioral studies in animais and psychophysical measurements in humans (cf. the reviews in Phillips, 1987, and Viemeister, 1988a, 1988b). However, several basic characteristics of auditory intensity perception still lack a unequivocal neurophysiological explanation, and even the question of whether their origin is to be found at the auditory periphery or in more central parts of the auditory system remains unanswered. This is the case in particular of die so-called "near-miss to" and "severe departure from" Weber's law. The near-miss to Weber's law corresponds to the fact that intensity discrimination thresholds for pure tones, when expressed as a fraction of stimulus intensity, decrease slightly as stimulus intensity increases, whereas Weber's law would require that they remain constant (McGill & Goldberg, 1968a, 1968b; Moore & Raab, 1974). The severe departure from Weber's law corresponds to an elevation in the Weber fraction at midintensities. The severe departure was mainly studied using brief high-frequency tone bursts (Carlyon & Moore, 1984, 1986a, 1986b). However, a similar increase in Weber functions at mid intensities was originally observed using clicks (Raab &

Stéphane Gallégo and Christophe Micheyl, Unite d'Enseignement et de Recherche Associee 5020, Centre National de la Recherche Scientifique, Pavillon U Hopital E. Herriot, Lyon, France.

This work was supported by a grant from the MXM Company. We are grateful to R. P. Carlyon and R. V. Shannon for helpful comments on an earlier version of this article. L. Collet and J. D. Durrant are acknowledged for valuable comments on die results of this study. M. Rolland is acknowledged for taking part in the measurements in normal-hearing listeners.

Correspondence concerning this article should be addressed to Christophe Micheyl, UPRESA CNRS 5020, Pavillon U Hopital E. Herriot, 69003 Lyon, France. Electronic mail may be sent to cmicheyl mucosa.univ-lyon I .fr.

Taub, 1969a, 1969b). More recently, tnidlevel humps in the function relating intensity discrimination thresholds to stimulus intensity were evidenced for tones in the condition of forward masking (Zeng & Turner, 1992; Zeng, Turner, & Relkin, 1991). Whether the origin of ail these phenomena is central or peripheral, and if it is peripheral, whether it relies on cochlear mechanisms or on auditory nerve (AN) functioning, is still a matter of debate. Models based on electrophysiological recordings of AN responses indicate that only a few AN fibers with characteristic frequencies akin to that of the signal are needed to account for the intensity-discrimination performances observed in humans, but such models do not clearly account for the decrease in Weber fraction at high intensities (Viemeister, 1988a, 1988b). Regarding the existence of a midlevel hump in Weber functions, the data obtained in unmasked conditions have generally been interpreted to support a peripheral origin (Carlyon & Moore, 1984, 1986a, 1986b; Long & Cullen, 1986; Raab & Taub, 1969a, 1969b); the effects obtained in nonsimultaneous masking conditions, although originally interpreted in terras of peripheral mechanisms (Zeng & Turner, 1992; Zeng et al., 1991), are now thought to reflect more central processes (Plack & Viemeister, 1992a, 1992b).

To gain further insight into the neurophysiological mechanisms underlying the encoding of sound intensity in die auditory system, a comparison of psychophysical and physiological data is required. However, in mort cases, such a comparison is rendered difficult by the fact that the neurophysiological and the psychophysical data to be compared corne from different studies and have often been obtained using different listeners or even species. During the two preceding decades another approach to the study of the neurophysiological basis of intensity perception has become available with the development of implanted auditory prostheses delivering direct electrical stimulation to neural stages of the auditory system in humans (Shannon, 1983; Shannon & Otto, 1990). Recent studies in listeners with such auditory implants have provided important information

about the neurophysiological origins of perceptual phenomena relating to the perception of intensity Nelson, Schmitz, Donaldson, Viemeister, & Javel, 1996; Zeng & Shannon, 1994, 1995).

In this study, we took advantage of the two approaches currently available to investigate the neurophysiological bases of intensity perception in humans: comparisons of physiological and psychophysical measures obtained in the same listeners and comparisons of physiological and psychophysical measures between listeners with normal auditory function and listeners with an implanted auditory prosthesis. Psychophysical and physiological measures were taken from listeners with normal hearing (NH) and listeners with a cochlear implant (CI). The physiological measurements consisted of auditory brainstem responses (ABRs). The psychophysical measurements consisted of intensity-discrimination limens (IDLs) measurements. All measurements were performed at different levels spanning almost the entire audible range of the listeners. Because we wanted to make comparisons between the two kinds of measurements and the two kinds of listeners, it was important to use similar stimuli throughout the study. To minimize the differences attributable to the respective specificities of the acoustic and electrical modes of stimulation, we used stimuli with a similar temporal structure (i.e., trains of electrical or acoustic pulses presented at the same rate and having the same overall duration).

Six CI listeners (6 men, aged 28-64 years) and 8 NB listeners (5 women and 3 men, aged 20-37 years) participated in the experiment. The CI listeners were implanted with a MXM DX10 multichannel electrode device (Beliaeff, Dubus, Leveau, Repetto, & Vincent, 1994). All suffered profound bilateral sensorineural hearing loss acquired postlingually 1-8 years before implantation. In NH listeners, pure-tone audiometry was performed using a Madsen OB 822 audiometer and TDH39 earphones to ensure that auditory thresholds were not larger than 10 dB hearing loss at octave frequencies between 125 and 8000 Hz.

ABR recordings were performed using a Nicolet Pathfinder II system. The positive, negative, and reference recording electrodes were placed on the forehead, the ear lobe ipsilateral to the side of stimulation, and the contralateral ear lobe, respectively. In CI listeners, the evoking stimuli consisted of 500-ms trains of biphasic pulses. The pulses were delivered to the implant device at a rate of 60 Hz using a dedicated stimulation system (MXM Digistim). The stimulation mode was common ground. To reduce long-term stimulation fatigue and to avoid the eventual occurrence of recurrent electrical interferences, the pulse trains were separated by a silent interval of 500 #177; 100 ms. ABRs .were successively recorded at 14 stimulus levels equally spaced on a linear scale within the dynamic range (i.e., the range between the detection

threshold and the loudest bearable level in the considered CI listener). Stimulus level was controlled by varying the charge per pulse phase. Pulse amplitude remained constant; only pulse duration was varied. In each patient, the electrode with the most apical location inside the cochlea was systematically tested. For normal insertion of the electrode array into the cochlea, the location of the most apical electrode corresponded roughly to the end of the first cochlear turn. Responses were bandpass-filtered between 0.2 and 8000 Hz before being acquired at a sampling rate of 50 kHz. The averaging process involved 3,072 sweeps. Subsequent processing--adaptive filtering--was thereafter performed in die digital domain to remove stimulus artifacts. Further details about die ABR recording and processing methods may be found in previous articles (Gallégo et al., 1996; Gallégo, Truy, Morgon, & Collet, 1997). In NH listeners, ABRs were evoked using acoustic clicks. The clicks were hi-pass-filtered at a cutoff frequency of 2000 Hz. This filtering reduced trie contribution from apical generators, known to be less well synchronized than most basal generators, and at the sanie time reduced spectral cues by ensuring that successive peaks in die long-terni spectrum of the stimulus would not be resolved. The equivalent rectangular bandwidth of the auditory filter around 2000 Hz was already about four times larger than die spacing between adjacent spectral peaks corresponding to the 60-Hz click repetition rate used. As in CI listeners, the stimulus trains had a duration of 500 ms and were separated by a silent interval of 500 #177; 100 ms. The clicks were delivered in the listeners' right ear through Sony CD450 headphones. Traces at each intensity were obtained by averaging 3,000 responses. Responses were analog filtered between 100 and 1500 Hz and were then sampled at a rate of 50 kHz before further digital processing performed to cancel stimulus artifacts. ABRs were recorded at 14 stimulus intensities in 6-dB steps from 6 to 84 dB above threshold. Participants reclined in a soundproof room.

The amplitude and latency of Waves I and III were collected in NH listeners. In CI listeners, because Wave I corresponds to an electric stimulus artifact, Waves II and III were considered. Wave I in NH listeners and Wave II in CI listeners are comparable in die sense that they reflect the earliest ABR components of neural origin and are likely to originate in the activity of auditory nerve fibers.

In NH listeners, intensity discrimination limens were measured using a two-interval, two-alternative, forced-choice (2A-2IFC) procedure with a two-down/one-up adaptive rule estimating the 71% correct point on the psychometric functions. In CI listeners, a 2A-2IFC procedure combined with a method of limits was used. In both procedures, the listener's task was to indicate which of two successive intervals contained the loudest stimulus. The stimuli used for IDL measurements were the same as those used for ABRs, namely electrical biphasic pulses and acoustic clicks presented at a rate of 60 Hz in CI and NH listeners, respectively. The stimuli were 500 ms long. The interstimulus interval was 500 ms. IDL measurements were performed at levels equal to those used in ABR recordings.

Figure 1 shows the average IDL function obtained from the 6 CI listeners. IDLs are expressed on the ordinate as

Figure 1. Intensity-discrimination limens (IDLs) as a function of stimulus level in cochlear implant listeners. The stimulus level is expressed as a percentage of the dynamic range on the abcissa. The mean IDLs across listeners are expressed as Weber fractions in decibels on the ordinate. The error bars represent the standard errors of the mean across listeners.

Weber fractions in decibels;' stimulus intensities are expressed on the abscissa as the percentages of dynamic range.² The DL function could be fitted by a line with a slope of --0.03 (R² = .73).

Figure 2 shows the average 1DL function measured in the 8 NH listeners. This function exhibited a clearly nonmono-

Figure 2. Intensity-discrimination limens (IDLs) as a function of stimulus level in normal-hearing listeners. The stimulus level is expressed in decibels SL on the abcissa. The mean IDLs across listeners are expressed as Weber fractions in decibels on the ordinate. The error bars represent the standard errors of the mean across listeners.

tonic shape characterized by a hump at intermediate stimulation levels. The statistical significance of this midlevel hump was indicated by a one-way repeated measures analysis of variance (ANOVA) performed on the data with the IDL as the dependent variable and the stimulus level as the independent variable, F(13, 91) = 2.02, p < .05. Post hoc comparisons performed using the Student's paired t test with a Bonferroni correction revealed that IDLs at 42 dB SL were significantly larger than IDLs at 18 dB SL, t(7) = 3.28, p = .013, and 72 dB SL, t (7 ) 8.06,p < .001.

Figure 3 shows the peak amplitudes of ABR Waves II and III as a function of stimulus level in CI listeners. ABR Wave

II and Wave III amplitude increased with stimulus level. Repeated measures ANOVAs indicated a significant effect of stimulus level on ABR Wave II and Wave III amplitudes: F(50, 10) = 8.39, p < .001, for Wave II; F(50, 10) = 6.43, p < .001, for Wave III. The monotonic growth functions of the two waves could be fitted with lines (slope = 0.14, R² = .96 for Wave II; slope = 0.23, R² = .98 for Wave III). Figure 4 shows the latencies of Waves II and III in the CI listeners. No significant dependency of Wave II amplitude on stimulus level was observed. The latency of Wave III decreased slightly but significantly, F(50, 10) = 2.13, p < .05, as stimulus level increased.

III in the NH listeners. Contrary to what was observed in the CI listeners, in the NH listeners Wave I amplitude did not significantly increase with stimulus level. A significant dependency of Wave III amplitude on stimulus level was indicated by a repeated measure ANOVA, F(77, 11) = 10.81, p < .001. However, the growth function relating Wave III amplitude to stimulus level stagnated over a large range of intensities from 24 to 54 dB SL before showing a sharp increase. This was confirmed statistically using a repeated measures ANOVA, which indicated no significant differences between ABR Wave III amplitudes for stimulus intensities between 12 and 54 dB SL, no significant differ-

Various ways of expressing intensity-discrimination thresholds have been used in die literature. We used Weber fractions in decibels in this study because this metric was used in most of the studies concemed primarily with die departures from Weber's law. Furthermore, its adequacy for analyzing data from cochlear implant listeners has recently been demonstrated (Nelson, Schmitz, Donald- son, Viemeister, & Javel, 1996).

² The specification of stimulus levels in terms of percentages of the dynamic range radier than in decibels SPL in cochlear implant listeners was motivated by results from the literature that suggest that die Weber fractions for electrical stimulation are a function of the level of the stimulus expressed in the percentage of dynamic range radier than in decibels SL (Nelson, Schmitz, Donaldson, Viemeister, & Javel, 1996).

GALLÉGO AND MICHEYL

I I

I I 1 I I

180

160-

0- - - I

0- 111

140-

120-

-o

100-

w
>

80-

60-

40-

20-

I I

1 1

1 1 1 1 1

60 70 80 90 100

0 6

12 18

60 66 72 78 84 90

Figure 3. Auditory brainstem response (ABR) Waves II and Wave III amplitudes as a function of stimulus level in cochlear implant listeners. The stimulus levels expressed as percentages of the dynamic range on the abcissa are the same as in Figure 1. The mean amplitude of each wave across listeners is expressed in nanovolts on the ordinate. The hollow symbols correspond to Wave II. The filled symbols correspond to Wave III. The error bars represent die standard errors of die mean across listeners.

ences between ABR Wave III amplitudes for stimulus intensifies between 60 and 84 dB SL, but statistically significant differences between ABR Wave III amplitudes overall, F(11, 77) = 10.81, p < .001. Post hoc mean

Figure 5. Auditory brainstem response (ABR) Waves I and Wave III amplitudes as a function of stimulus level in normal-hearing listeners. The stimulus levels expressed in decibels SL on the abcissa are the sanie as in Figure 2. The mean amplitude of each wave across listeners is expressed in nanovolts on the ordinate. The hollow symbols correspond to Wave I. The filled symbols correspond to Wave III. The error bars represent die standard errors of die mean across listeners.

comparisons evidenced a significant difference between ABR Wave III amplitudes at 54 and 60 dB SL (Student's paired t test), t(7) = 4.94, p < .005.

Figure 4. Auditory brainstem response (ABR) Waves II and Wave III latencies as a function of stimulus level in cochlear implant listeners. The stimulus levels expressed as percentages of die dynamic range on the abcissa are die sanie as in Figures 1 and 3. Wave II mean latencies across listeners are indicated in milliseconds on the left ordinate. Wave III latencies are indicated on the right ordinate. The hollow symbols correspond to Wave II. The filled symbols correspond to Wave III. The error bars represent the standard errors of die mean across listeners.

Figure 6. Auditory brainstem response (ABR) Waves I and Wave III latencies as a function of stimulus level in normal-hearing listeners. The stimulus levels expressed in decibels SL on the abcissa are die same as in Figures 2 and 5. Wave I mean latencies across listeners are indicated in milliseconds on die left ordinate. Wave III latencies are indicated on die right ordinate. The hollow symbols correspond to Wave I. The filled symbols correspond to Wave III. The error bars represent die standard errors of the mean across listeners.

listeners. The latencies of the two waves both decreased highly significantly with increasing stimulus level: F(77, 11) = 25.02, p < .001, for Wave I; F(77, 11) = 14.75, p < .001 for Wave III.

The main finding of this study came from the comparison of IDL functions between NH and CI listeners: Whereas in CI listeners IDL decreased overall slightly with increasing intensity, in NH listeners IDLs were found to be larger at midintensities. A second finding came from the comparison of the ABR amplitude functions between NH and CI listeners: Whereas in CI listeners the amplitude of ABR Waves II and III showed a grossly monotonie and regular dependence on stimulus level, in NH listeners ABR Waves I and III amplitudes did not significantly increase as a function of stimulus level over a large range of levels, although the latencies of the corresponding waves did significantly vary.

The observation that in CI listeners IDLs, expressed as Weber fractions, decreased when stimulus intensity was increased over the whole dynamic range is in overall agreement with results from several previous studies on intensity discrimination with electric stimulation in humans (Nelson et al., 1996; Shannon, 1983, 1992). Such a decrease in Weber fractions with increasing stimulus intensity was initially observed in NH listeners and dubbed a near-miss to Weber's law (McGill & Goldberg, 1968a, 1968b). The origin of the near-miss to Weber's law is unclear. One hypothesis is that it cornes from a broadening in the spread of excitation elicited in the peripheral auditory system as intensity is increased (Viemeister, 1983, 1988a, 1988b). In NH listeners, this level-dependent increase in the spread of excitation would occur because of nonlinear cochlear mechanisms. In CI listeners, although cochlear mechanisms are bypassed by direct electrical stimulation of the AN, some form of nonlinear spread of excitation might still occur because of spread of electrical current into the spiral ganglion (Nelson et al., 1996). In this respect, it is noteworthy that the ABR results obtained in our study, in agreement with the results from previous studies (Abbas & Brown, 1991), did not show the increase in the rate of growth of Wave II amplitude that might be expected if the rate of recruitment of AN fibers became larger at higher stimulus levels. However, because the measured peak amplitude depended not only on the number of units firing but also on the synchrony of the underlying neural events (Goldstein & Kiang, 1958), one cannot rule out the possibility that the expected increase in the rate of recruitment with increasing stimulus intensity was compensated by a lower increase in synchronization across units when more units were recruited.

A second important result of our study was the "midlevel hump" in the Weber function in NH but not in CI listeners. An increase in Weber fractions at midlevels comparable to the one observed in this study has been reported in previous studies on intensity discrimination for clicks (Avakyan & Radianova, 1963; Raab & Taub, 1969a).³ This effect,

however, failed to be replicated in a later study (Penner & Viemeister, 1973). In more recent studies, even larger increases in Weber fractions at midstimulation levels were found using short-duration, high-frequency tone bursts (Carlyon & Moore, 1984, 1986a, 1986b; Long & Cullen, 1986). Ultimately, large midlevel humps in intensity discrimination were observed with tone bursts in the condition of forward masking (Zeng et al., 1991). The origin of the increases in intensity discrimination thresholds observed in these varions studies remains unclear, and whether these effects share the same underlying mechanisms is uncertain. The psychophysical measures of IDLs obtained by Raab and Taub (1969a) were compared with recordings of the AN compound action potential in a companion study (Raab & Taub, 1969b). The results indicated that the effect was already present at the level of the AN, suggesting a peripheral origin. The studies by Carlyon and Moore (1984, 1986a, 1986b) brought forward several arguments suggesting that the midlevel hump observed with brief-duration, high-frequency tons originated at the auditory periphery. A later study of this effect further suggested that it could reflect the functioning of cochlear micromechanisms (Long & Cullen, 1986). On the contrary, the midlevel humps in intensity discrimination observed in nonsimultaneous masking conditions (Zeng & Turner, 1992; Zeng et al., 1991) were suggested to have a central origin (Plack & Viemeister, 1992a, 1992b; Zeng & Shannon, 1995). However, the relation of such effects to the effects obtained in unmasked conditions remain uncertain, and interpretations relative to their origin appear less relevant in the framework of this study. Therefore, to summarize, most, if not all, the data available in the literature on midlevel humps in unmasked conditions point to a peripheral origin.

The results of this study bring a new argument for this hypothesis. The observation of a midlevel hump in DL functions in NH listeners and the absence of this effect in CI listeners indeed suggests that the origin of this effect is to be

³ The fact that the size of the effect obtained in our study with click trains was comparable to that found in the previous study using clicks in isolation (Raab & Taub, 1969a) is noteworthy. This similarity of results suggests that the midlevel hump observed with isolated clicks was not overridden by integrative processes operating in the time domain when the stimulus was iterated. This result appears to be at variance with the hypothesis--which might be put forward to account for the fact that midlevel humps in intensitydiscrimination limens functions have generally been reported with transient or short-duration stimuli but not with longer stimuli (Carlyon & Moore, 1984)--that integrative processes operating with long-duration stimuli may compensate for the paucity of both spectral and temporal information present in short-duration stimuli, broad-bandwidth stimuli, or both. Nevertheless, the low stimulus rate used in our study leaves open the possibility that a multiple- looks mechanism, although possibly operating at higher rates, was reset from one click in the train to the other. It has been shown that clicks in a pair were integrated only when they were separated by less than about 10 ms (Viemeister & Wakefield, 1991), which is well below the 16.66-ms interval corresponding to the 60-Hz click train used in our study.

⁴ The difference observed between the ranges of auditory brainstem response (ABR) wave amplitudes in cochlear implant (CI) and normal-hearing (NH) listeners may appear as a limitation to the validity of direct comparisons between the ABR growth functions obtained in the two types of listeners. However, note that the comparisons made in this study did not regard the absolute amplitude of these waves but the way they varied as a function of stimulus intensity throughout the perceptual dynamic range. The intensities used to evoke ABRs and measure Weber fractions spanned the perceptual dynamic range similarly in CI and NH listeners. The largest amplitudes obtained in CI listeners were likely to have been caused by the fact that direct electric stimulation of the AN elicited more synchronous discharges across AN fibers than acoustic stimulation (Hartmann, Topp, & Klinke, 1984; Javel, 1989; Parkins, 1989). It is not clear whether discharge synchrony across fibers is a factor of perceived intensity.

Abbas, P. J., & Brown, C. J. (1991). Electrically evoked auditory brainstem response: Growth of response with current level. Hearing Research, 51, 123-137.

Avakyan, R. V., & Radianova, E. A. (1963). The special features of differential intensity thresholds for a brief sound signal. Sovietic Physical Acoustics, 8, 320-323.

Beliaeff, M., Dubus, R, Leveau, J. M., Repetto, J. C., & Vincent, P. (1994). Sound signal processing and stimulation coding of the Digisonic DX10 15-channel cochlear implant. In I. J. HochmairDesoyer & E. S. Hochmair (Eds.), Advances in cochlear implants (pp. 198-203). Vienna, Austria: Manz.

Carlyon, R. R, & Moore, B. C. J. (1984). Intensity discrimination: A "severe departure" from Weber's law. Journal of the Acoustical Society ofAmerica, 76, 1369-1376.

Carlyon, R. P., & Moore, B. C. J. (1986a). Continuons versus gated pedestals and the "severe departure" from Weber's law. Journal of the Acoustical Society ofAmerica, 79, 453-460.

Carlyon, R. P., & Moore, B. C. J. (1986b). Detection of tones in noise and the "severe departure" from Weber's law. Journal of the Acoustical Society ofAmerica, 79, 461-464.

Delgutte, B. (1987). Peripheral processing of speech information: Implications from a physiological study of intensity discrimination. In M. E. H. Schouten (Ed.), Psychophysics and speech perception (pp. 333-353). Dordrecht, The Netherlands: Nijhoff.

Eggermont, J. J., & Don, M. (1980). Analysis of the click-evoked brainstem potentials in humans using high-pass noise masking: II. Effect of click intensity. Journal of the Acoustical Society of America, 68, 1671-1675.

Gallégo, S., Micheyl, C., Berger-Vachon, C., Truy, E., Morgon, A., & Collet, L. (1996). Ipsilateral ABR with cochlear implant. Acta Otolaryngologica (Stockholm), 116, 228-233.

Gallégo, S., Truy, E., Morgon, A., & Collet, L. (1997). EABRs and surface potentials with a transcutaneous multielectrode cochlear

implant. Acta Otolaryngologica (Stockholm), 117, 164-168. Goldstein, M. H. Jr., & Kiang, N. Y. S. (1958). Synchrony of neural activity in electric responses evoked by transient acoustic

found at the level of peripheral auditory processes that are present in NH listeners and bypassed in CI listeners, namely cochlear mechanics and hair-cell transduction processes. A possible interpretation of the relationship between the worsening of intensity-discrimination performances at midintensities and the functioning of the cochlea cornes from indications that the amplitude of basilar membrane motion grows more slowly with acoustic stimulus amplitude at moderate than at low and high levels of stimulation (Robles, Ruggero, & Rich, 1986; Yates, 1990). If stimuli are discriminated in intensity on the basis of the difference in the discharge rates they evoke in auditory neurons, as suggested by several studies (Delgutte, 1987; Viemeister, 1983, 1988a, 1988b; Winslow & Sachs, 1988), the degraded intensitydiscrimination performances observed at midievels in the present and previous studies might be explainable in terms of cochlear mechanisms, as previously suggested (Long & Cullen, 1986). That would also explain that in CI listeners, cochlear mechanisms being bypassed by direct electrical stimulation of the AN, no such worsening in intensitydiscrimination performance was observed.

Another argument for a peripheral origin of the midlevel hump, which may tentatively be brought forward on the basis of the physiological results of our study, cornes from the observation of a difference between the ABR wave growth functions in NH and CI listeners. That is, whereas the amplitude of ABR Waves II and III increased regularly with stimulus level in CI listeners, in NH listeners no significant increase in the amplitude of the ABR waves was observed over a large range of moderate intensities.⁴ This stagnation in ABR amplitude appears to correspond to the initial saturation reported in a previous study on the dependence of ABR amplitude on stimulus level in NH listeners (Eggermont & Don, 1980). The exaggerated flatness of the growth functions shown here was attributable to the fact that near-zero ABR amplitudes at die lowest stimulus levels were not displayed and to the use of a linear y-axis scale radier than a logarithmic one, which would have expanded the rapid amplitude increase at low stimulus levels. The absence of a significant variation in ABR amplitude at moderate intensities in NH listeners could not simply be attributed to insufficient measurement sensitivity because marked variations in the latency of die same ABR waves over a corresponding range of stimulus levels were observed. Under the hypothesis that differences in stimulus intensity are encoded in the auditory system as differences in overall neural activity, reduced variations in compound neural responses should be related to reduced intensity discrimination performances. In this view, die reduced dependency of ABR amplitudes on stimulus level observed in NH listeners as compared with CI listeners may be tentatively related to the observation of reduced intensity-discrimination performance at moderate levels in these NH listeners. However, because no clear quantitative relationship between the slope of the ABR functions and the intensity discrimination thresholds could be established in our study, neither at the

Hartmann, R., Topp, G., & Klinke, R. (1984). Discharge patterns of cat primary auditory fibers with electrical stimulation of the cochlea. Hearing Research, 13, 47-62.

Javel, E. (1989). Acoustic and electrical encoding of temporal information. In J. Miller & F. A. Spelman (Eds.), Cochlear implants: Models of the electrically stimulated ear (pp. 247-

Long, G. R., & Cullen, J. K. Jr. (1986). Intensity difference limens at high frequencies. Journal of the Acoustical Society ofAmerica, 78, 507-513.

McGill, W. J., & Goldberg, J. P. (1968a). A study of the near-miss involving Weber's law and pure-tone intensity discrimination.

McGill, W. J., & Goldberg, J. P. (1968b). Pure-tone intensity discrimination and energy detection. Journal of the Acoustical Society ofAmerica, 44, 576-581.

Moore, B. C. J., & Raab, D. H. (1974). Pure-tone intensity discrimination: Some experiments relating to the near-miss to

Nelson, D. A., Schmitz, J. L., Donaldson, G. S., Viemeister, N. F., & Javel, E. (1996). Intensity discrimination as a function of stimulus level with electric stimulation. Journal of the Acoustical Society of America, 100, 2393-2414.

Parkins, C. W. (1989). Temporal response patterns of AN fibers to electrical stimulation in deafened squirrel monkeys. Hearing Research, 41, 137-168.

Penner, M. J., & Viemeister, N. F. (1973). Intensity discrimination: The effects of click bandwidth and background noise. Journal of the Acoustical Society of America, 54, 1184-1188.

Phillips, D. P. (1987). Stimulus intensity and loudness recruitment: Neural correlates. Journal of the Acoustical Society of America, 82, 1-12.

Plack, C., & Viemeister, N. F. (1992a). The effects of notched noise on intensity discrimination under forward masking. Journal of the Acoustical Society of America, 92, 1902-1910.

Plack, C., & Viemeister, N. F. (1992b). Intensity discrimination under backward masking. Journal of the Acoustical Society of America, 92, 3097-3101.

Raab, D. H., & Taub, H. A. (1969a). Click-intensity discrimination with and without background masking noise. Journal of the Acoustical Society ofAmerica, 46, 965-968.

Raab, D. H., & Taub, H. A. (1969b). Fluctuations in N1 amplitude in relation to click-intensity discrimination. Journal of the Acoustical Society ofAmerica, 46, 969-978.

Shannon, R. V. (1983). Multichannel electrical stimulation of the auditory nerve in man: I. Basic psychophysics. Hearing Research, 11, 157-189.

Shannon, R. V. (1992). Temporal modulation transfer function in patients with cochlear implants. Journal of the Acoustical Society of America, 91, 2156-2163.

Shannon, R. V., & Otto, S. R. (1990). Psychophysical measures from electrical stimulation of the human cochlear nucleus. Hearing Research, 47, 159-168.

frequencies in the presence of noise. Science, 221, 1206-1208.
Viemeister, N. E (1988a). Intensity coding and the dynamic range

Viemeister, N. F. (1988b). Psychophysical aspects of auditory intensity coding. In G. Edelman, W. E. Gall, & W. M. Cowan (Eds.), Auditory function (pp. 213-241). New York: Wiley.

Viemeister, N. F., & Wakefield, G. H. (1991). Temporal integration and multiple looks Journal of the Acoustical Society ofAmerica, 90, 858-865.

Winslow, R. L., & Sachs, M. B. (1988). Single-tone intensity discrimination based on AN rate responses in backgrounds of quiet, noise, and with stimulation of the crossed olivocochlear bundle. Hearing Research, 35, 165-190.

Yates, G. K. (1990). The basilar membrane input-output function. In W. L. Madison, R Dallos, C. D. Geisler, J. M. Matthews, M. A. Ruggero, & C. R. Steele (Eds.), The mechanics and biophysics of hearing (pp. 106-113). New York: SpringerVerlag.

Zeng, E-G., & Shannon, R. V. (1994). Loudness coding mechanisms inferred from electric stimulation of the human auditory system. Science, 264, 564-566.

Zeng, F.-G., & Shannon, R. V. (1995). Possible origins for the non-monotonic intensity discrimination function in forward masking. Hearing Research, 82, 216-224.

Zeng, E-G., & Turner, C. W. (1992). Intensity discrimination in forward masking. Journal of the Acoustical Society of America, 92, 782-787.

Zeng, E-G., Turner, C. W., & Relkin, E. M. (1991). Recovery from prior stimulation: H. Effects upon intensity discrimination. Hearing Research, 55, 223-230.

Received September 2, 1997 Revision received January 9, 1998 Accepted January 28, 1998

Des travaux (Hermann et al, 1992 ; Gallégo et a1,1997, Groenen et al, 1997) ont montré qu'il était possible d'estimer une partie de la reconnaissance des sujets implantés cochléaires a partir de mesures électro-ph ysiolog igues.

Nous avons voulu évaluer la possible relation qu'il existe entre les PEAEP et le reconnaissance sans lecture labiale sur une population de 17 implantés cochléaires.

Contrainte Psycho-Physiques et Electrophysiologiques sur le codage de la stimulation électrique chez les sujets porteurs d'un implant cochléaire

- Article 19 :

INTENSITY DISCRIMINATION AND AUDITORY BRAINSTEM RESPONSES IN COCHLEAR
IMPLANT AND NORMAL-HEARING LISTENERS

Contrainte Psycho-Physiques et Electrophysiologiques sur le codage de la stimulation électrique chez les sujets porteurs d'un implant cochléaire

- Article 19 :

INTENSITY DISCRIMINATION AND AUDITORY BRAINSTEM RESPONSES IN COCHLEARIMPLANT AND NORMAL-HEARING LISTENERS

INTENSITY DISCRIMINATION AND AUDITORY BRAINSTEM RESPONSES IN COCHLEAR
IMPLANT AND NORMAL-HEARING LISTENERS