Responses of "high-spontaneous" auditory-nerve fibers to consonant-vowel syllables in noise

Responses of "high-spontaneous" single auditory-nerve fibers in anesthetized cat to nine different spoken stop and nasal consonant-vowel syllables presented in four different levels of speech-shaped noise are reported. The temporal information contained in the responses was analyzed using "composite" spectrograms and pseudo-3D spatial-frequency plots. Spectral characteristics of both consonant and vowel segments of the CV syllables were strongly encoded at S/N ratios of 30 and 20 dB. At S/N = 10 dB, formant information during the vowel segments was all that was reliably detectable in most cases. Even at S/N = 0 dB, most vowel formants were detectable, but only with relatively long analysis windows (40 ms). The increases (and decreases) in discharge rate during various phases of the responses were also determined. The rate responses to the "release" and to the voicing of the stop-consonant syllables were quite robust, being detectable at least half of the time, even at the highest noise level. Comparisons with psychoacoustic studies using similar stimuli are made.     

A temporal analysis of auditory-nerve fiber responses to spoken stop consonant-vowel syllables

Auditory-nerve fiber spike trains were recorded in response to spoken English stop consonant-vowel syllables, both voiced (/b,d,g/) and unvoiced (/p,t,k/), in the initial position of syllables with the vowels /i,a,u/. Temporal properties of the neural responses and stimulus spectra are displayed in a spectrographic format. The responses were categorized in terms of the fibers' characteristic frequencies (CF) and spontaneous rates (SR). High-CF, high-SR fibers generally synchronize to formants throughout the syllables. High-CF, low/medium-SR fibers may also synchronize to formants; however, during the voicing, there may be sufficient low-frequency energy present to suppress a fiber's synchronized response to a formant near its CF. Low-CF fibers, from both SR groups, synchronize to energy associated with voicing. Several proposed acoustic correlates to perceptual features of stop consonant-vowel syllables, including the initial spectrum, formant transitions, and voice-onset time, are represented in the temporal properties of auditory-nerve fiber responses. Nonlinear suppression affects the temporal features of the responses, particularly those of low/medium-spontaneous-rate fibers.     

Responses of auditory-nerve fibers to multiple-tone complexes

To relate level-dependent properties of auditory-nerve-fiber responses to nasal consonant-vowels to the basic frequency selective and suppressive properties of the fibers, multitone complexes, with the amplitude of a single (probe) component incremented, were used as stimuli. Quantitative relations were obtained between the systematic increase of fiber synchrony to the probe tone and the decrease of synchrony to CF, as the amplitude of the probe tone was increased. When such relations are interpreted as a measure of fiber frequency selectivity based on a relative synchrony criterion, a breadth of frequency tuning is obtained, at a 70-dB SPL multitone sound-pressure level, which is generally broader than that of the fiber's threshold tuning curve. Quantitative comparisons with the same fiber's responses to the nasal speech sounds indicate that the fiber's speech responses share some common features with its probe-tone responses.     

Responses of "lower-spontaneous-rate" auditory-nerve fibers to speech syllables presented in noise. II: Glottal-pulse periodicities.

The responses of single cat auditory-nerve fibers to naturally spoken voiced sounds (the vowels [a, i, u] and the murmur [m]) presented at normal intensity (70 dB SPL) in various levels of speech-shaped noise were analyzed for the encoding of the glottal-pulse (fundamental) period. To quantify the strength of this fundamental-period encoding, selected segments of the response histograms were autocorrelated, rectified, and fitted with the best-fitting sinusoid of the fundamental frequency. The magnitude of this best-fitting sinusoid was taken as the magnitude of synchronization. In most cases, it was found that the "lower-SR" fibers (those with spontaneous discharge rates less than 20/s) encoded the fundamental periodicity more strongly and more robustly than did the "high-SR" fibers (those with spontaneous discharge rates greater than 20/s). When either a single strong spectral peak or a relatively "flat" spectrum excited a fiber, it showed poor synchronization to the fundamental period, regardless of its spontaneous-rate class. Judging from a few examples, the glottal-pulse synchronization appears to be intensity dependent, with the relative performance of the high-SR fibers improving at lower intensities. A conceptual model is given which accounts for the general characteristics of the data.     

Responses of "lower-spontaneous-rate" auditory-nerve fibers to speech syllables presented in noise. I: General characteristics.

Responses of auditory-nerve fibers in anesthetized cats to nine different spoken stop- and nasal-consonant/vowel syllables presented at 70 dB SPL in various levels of speech-shaped noise [signal-to-noise (S/N) ratios of 30, 20, 10, and 0 dB] are reported. The temporal aspects of speech encoding were analyzed using spectrograms. The responses of the "lower-spontaneous-rate" fibers (less than 20/s) were found to be more limited than those of the high-spontaneous-rate fibers. The lower-spontaneous-rate fibers did not encode noise-only portions of the stimulus at the lowest noise level (S/N = 30 dB) and only responded to the consonant if there was a formant or major spectral peak near its characteristic frequency. The fibers' responses at the higher noise levels were compared to those obtained at the lowest noise level using the covariance as a quantitative measure of signal degradation. The lower-spontaneous-rate fibers were found to preserve more of their initial temporal encoding than high-spontaneous-rate fibers of the same characteristic frequency. The auditory-nerve fibers' responses were also analyzed for rate-place encoding of the stimuli. The results are similar to those found for temporal encoding.     

The responses of models of "high-spontaneous" auditory-nerve fibers in a damaged cochlea to speech syllables in noise

The responses of four high-spontaneous fibers from a damaged cat cochlea responding to naturally uttered consonant-vowel (CV) syllables [m], [p], and [t], each with [a], [i], and [u] in four different levels of noise were simulated using a two-stage computer model. At the lowest noise level [+30 dB signal-to-noise (S/N) ratio], the responses of the models of the three fibers from a heavily damaged portion of the cochlea [characteristic frequencies (CFs) from 1.6 to 2.14 kHz] showed quite different response patterns from those of fibers in normal cochleas: There was little response to the noise alone, the consonant portions of the syllables evoked small-amplitude wide-bandwidth complexes, and the vowel-segment response synchrony was often masked by low-frequency components, especially the first formant. At the next level of noise (S/N = 20 dB), spectral information regarding the murmur segments of the [m] syllables was essentially lost. At the highest noise levels used (S/N = +10 and 0 dB), the noise was almost totally disruptive of coding of the spectral peaks of the consonant portions of the stop CVs. Possible implications of the results with regard to the understanding of speech by hearing-impaired listeners are discussed.     

Temporal resolution in frog auditory-nerve fibers

Sinusoidally amplitude-modulated (SAM) noise was monaurally presented to the neotropical frog, Eleutherodactylus coqui, while recording intracellularly from auditory-nerve fibers. Neuronal phase locking was measured to the SAM noise envelope in the form of a period histogram. The modulation depth was changed (in 10% steps) until the threshold modulation depth was determined. This was repeated for various modulation frequencies (20-1200 Hz) and different levels of SAM noise (34-64 dB/Hz). From these data, temporal modulation transfer functions (TMTFs) were produced and minimum integration time (MIT) for each auditory fiber was calculated. The median MIT was 0.42 ms (lower quartile 0.32, upper quartile 0.68 ms). A noise level-dependent effect was noted on the shape of the TMTF as well as the minimum integration time. The latter results may be explained as a loss in spectral resolution with increasing noise level, which is consistent with the correlation that was found between minimum integration time and bandwidth.     

Frequency glides in the impulse responses of auditory-nerve fibers

Previous reports of frequency modulations, or glides, in the impulse responses of the auditory periphery have been limited to analyses of basilar-membrane measurements and responses of auditory-nerve (AN) fibers with best frequencies (BFs) greater than 1.7 kHz. These glides increased in frequency as a function of time. In this study, the instantaneous frequency as a function of time was measured for impulse responses of AN fibers in the cat with a range of BFs (250-4500 Hz). Impulse responses were estimated from responses to wideband noise using the reverse-correlation technique. The impulse responses had increasing frequency glides for fibers with BFs greater than 1500 Hz, nearly constant frequency as a function of time of BFs between 750 and 1500 Hz, and decreasing frequency glides for BFs below 750 Hz. Over the levels tested, the glides for fibers at all BFs were nearly independent of stimulus level, consistent with previous reports of impulse responses of the basilar membrane and AN fibers. Implications of the different glide directions observed for different BFs are discussed, specifically in relation to models for the auditory periphery as well as for the derivation of impulse responses for the human auditory periphery based on psychophysical measurements.     

Effects of stimulus frequency on adaptation in auditory-nerve fibers.

Several experimental methods of depressing the response of auditory-nerve fibers to tonal stimuli have been shown to reduce the response to signal frequencies at or near fiber CF (characteristic frequency) more than to frequencies greater or less than CF. In this study we have used a short adapting tone presented before each test-tone burst to reduce the fiber's response to the test tone. We observed the effects of changing test frequency when the adapting frequency was held constant at fiber CF. The depression in discharge rate was found to be approximately constant across test frequency.     

Physiological mechanisms of psychophysical masking: observations from auditory-nerve fibers

Masking might be due either to the spread of the excitation produced by the masker to the place of the tone signal along the cochlea or to the suppression of the response to the signal by the masker. In order to identify the contributions of these two mechanisms to tone-on-tone masking, masked thresholds of auditory-nerve fibers were measured in anesthetized cats using the same stimulus paradigms and detection criteria as in psychophysics. Suppressive masking was identified by comparing thresholds for simultaneous masking with those for a nonsimultaneous masking technique resembling pulsation thresholds. These nonsimultaneous thresholds do not include the contribution of suppression to masking because suppression only occurs for stimuli that overlap in time. For each masker and signal frequency, the fibers with the lowest (or "best") masked thresholds had characteristic frequencies (CF) slightly on the opposite side of the masker frequency with respect to the signal frequency, consistent with the psychophysical phenomenon of off-frequency listening. Patterns of best masked thresholds against signal frequency resembled psychophysical masking patterns in that they showed a maximum for signal frequencies close to the masker, and a skew toward high frequencies. Masking was found to be both excitatory and suppressive, with the relative contribution of the two mechanisms depending on the frequency separation between signal and masker. Suppressive masking was large for signal frequencies well above the masker. For these conditions, simultaneous thresholds grew more rapidly with masker level than did nonsimultaneous thresholds, suggesting that the upward spread of masking is largely due to the growth of suppression rather than to that of excitation.     

Spectral characteristics and synchrony in primary auditory-nerve fibers in response to pure-tone acoustic stimuli

Under pure-tone stimulation, the spectrum of the period histogram recorded from primary auditory-nerve fibers at low and medium frequencies contains components at DC, at the applied tone frequency (the fundamental), and at a small number of harmonics of the tone frequency. The magnitudes and phases of these spectral components are examined. The spectral magnitudes of the fundamental and various harmonic components generally bear a fixed proportionality to each other over a broad range of signal conditions and nerve-fiber characteristics. This implies that the shape of the underlying rectified wave remains essentially unchanged over a broad range of stimulus intensities. For high-frequency stimuli, the fundamental and harmonic components are substantially attenuated. We provide a theoretical basis for the decrease of the spectralcomponent magnitudes with increasing harmonic number. For low-frequency pure-tone signals, the decrease is caused principally by the uncertainty in the position of neural-event occurrences within the half-wave-rectified period histogram. The lower the stimulus frequency, the greater this time uncertainty and therefore the lower the frequency at which the spectral components begin to diminish. For high-frequency pure-tone signals, on the other hand, the decrease is caused principally by the frequency rolloff associated with nervespike time jitter (it is then called loss of phase locking or loss of synchrony). Since some of this jitter arises from noise in the auditory nerve, it can be minimized by using peak detection rather than level detection. Using a specially designed microcomputer that measures the times at which the peaks of the action potentials occur, we have demonstrated the presence of phase locking to tone frequencies as high as 18 kHz. The traditional view that phase locking is always lost above 6 kHz is clearly not valid. This indicates that the placeversus-periodicity dichotomy in auditory theory requires reexaraination.     

Noise susceptibility and immunity of phase locking in amphibian auditory-nerve fibers

Recordings from auditory-nerve fibers in the anesthetized frog revealed that addition of broadband noise results in a reduction in the ability of a fiber to phase lock to a continuous pure tone. In particular, our results suggest that: (i) there is a threshold below which masking noise has little or no effect on vector strength (VS); then with increasing masking noise level, VS appears to decrease monotonically for all test frequencies (TFs); (ii) there exist subpopulations of auditory-nerve fibers in the frog for which the deterioration of phase locking to tones in wideband noise depends critically on the relationship of the TF to the fiber's CF. Specifically, in one subpopulation (43% of the fibers studied), the rate of VS decrease with increasing levels of masking noise is greater for CF tones than it is for TFs greater than CF. The net result is a "crossing" of the VS versus masking noise functions (e.g., Fig. 6); (iii) there exists a small subpopulation of amphibian papillar (a.p.) fibers for which the rate of VS decrease with increasing levels of masking noise is less for TFs less than CF than it is for CF tones (e.g., Fig. 5); (iv) there is a pronounced noise-induced phase lead for TFs greater than CF, whereas, for stimulus tones at or below CF, the preferred firing phase is nearly noise-level independent; (v) the remainder of the sample consists of fibers in which the VS-falloff rates appear to be test-frequency independent; (vi) addition of wideband masking noise to a CF tone, and increasing the CF-tone level in the absence of noise, produced (qualitatively) similar effects on the preferred firing phase of auditory-nerve fibers (e.g., Figs. 1 and 7). Thus amphibian auditory-nerve fibers appear to be energy detectors, i.e., exhibit phase shifts corresponding to the total energy within the filter passband defined by the frequency-threshold curve.     

A population study of auditory-nerve fibers in unanesthetized decerebrate cats: response to pure tones

Population responses of single auditory-nerve (AN) fibers to 1-kHz pure tones were investigated in unanesthetized decerebrate cats. Driven rate (spikes/s), Fourier component amplitude (spikes/s), and d'e = (driven rate)/(standard deviation) were examined. These results extend the presently available information. First, the AN population responses, previously observed only in anesthetized animals, are now observed in unanesthetized animals. The anesthetized and unanesthetized animal responses are qualitatively similar. Second, a detectability measure d'e was applied, à la signal detection theory, to populations of AN fibers. A previously unknown behavior of the low-spontaneous-rate (SR) (less than 15 spikes/s) fibers was observed: the d'e increased to large values at moderate and high stimulus levels, giving rise to a large, sharply tuned peak in the d'e-place profile, even when the driven rate was saturated. This behavior was absent in the high-SR (greater than 15 spikes/s) fibers. This observation suggests that the d'e of the low-SR fibers is the best code for stimulus intensity among the different response measures examined, and that the d'e-place profile may be a more precise spatial code for stimulus frequency than the commonly considered rate-place code.     

Sensitivity of auditory-nerve fibers to changes in intensity: a dichotomy between decrements and increments

Adaptation of auditory-nerve responses was investigated by applying increments and decrements in intensity to an ongoing tonal background. The change in firing rate produced by a change in intensity was obtained as a function of the time delay from the onset of the background to the onset of the change in intensity. The initial change in firing rate was measured using both small (1 ms) and large (10 ms) time intervals in order to evaluate properties of rapid and short-term adaptation, respectively. Consistent with previous results, the incremental and decremental responses measured with large windows were independent of time delay and the amount of prior adaptation. A similar additivity was observed for the incremental response measured with a small time window. In contrast, the decremental response measured with a small window decreased with increasing time delay and in proportion to the decrease in firing rate produced by the background. A similar decrease was observed in the response modulation produced by sinusoidal amplitude modulation. It was concluded that sensitivity to decrements in intensity decreases during adaptation, so that this response component does not reflect the additivity inherent in other aspects of adaptation.     

Response of auditory-nerve fibers to intensity increments in a multitone complex: neural correlates of profile analysis

Recent psychophysical studies have shown that the detection of an intensity increment superimposed on the center component (1 kHz) of a multitone complex (1, 3, 7, or 11 components) improves as more components are added outside of the critical band. It has been suggested that this form of intensity discrimination is based on a change in the neural profile. To test this hypothesis, neural profiles were constructed by plotting the degree of phase locking to the 1-kHz tone as a function of each unit's characteristic frequency (CF). Neural phase-locking profile to the 1-component signal at 1 kHz had a broad peak; however, the neural profile became narrower as the number of components in the signal increased. The just detectable increment for the 1-component condition was -5 dB re: 1000-Hz component level (3.86-dB increment plus component level re: component level), whereas, for the 3-, 7-, and 11-component conditions, it was -15 dB re: component level (1.42 dB). The neural and psychophysical IDL for the chinchilla were similar for the 1-component condition. However, the overall trends in the psychophysical and neural data are different. In the psychophysical studies IDL is typically poorest in the 3-component condition and improves when more components are added. By contrast, the neural IDL was poorest in the 1-component condition and improved when more components were added. In the multicomponent conditions, units with CFs in 492-1380 Hz were found to be more sensitive in detecting the intensity increment to the 1000-Hz component.     

Adaptation, saturation, and physiological masking in single auditory-nerve fibers.

Results are reviewed concerning some effects, at a units's characteristic frequency, of a short-term conditioning stimulus on the responses to perstimulatory and poststimulatory test tones. A phenomenological equation is developed from the poststimulatory results and shown to be consistent with the perstimulatory results. According to the results and equation, the response to a test tone equals the unconditioned or unadapted response minus the decrement produced by adaptation to the conditioning tone. Furthermore, the decrement is proportional to the driven response to the conditioning tone and does not depend on sound intensity per se. The equation has a simple interpretation in terms of two processes in cascade--a static saturating nonlinearity followed by additive adaptation. Results are presented to show that this functional model is sufficient to account for the "physiological masking" produced by wide-band backgrounds. According to this interpretation, a sufficiently intense background produces saturation. Consequently, a superimposed test tone cause no change in response. In addition, when the onset of the background precedes the onset of the test tone, the total firing rate is reduced by adaptation. Evidence is reviewed concerning the possible correspondence between the variables in the model and intracellular events in the auditory periphery.