First-spike timing of auditory-nerve fibers and comparison with auditory cortex

Information content of auditory cortical responses to time-varying acoustic stimuli

The present study explores the issue of cortical coding by spike count and timing using statistical and information theoretic methods. We have shown in previous studies that neurons in the auditory cortex of awake primates have an abundance of sustained discharges that could represent time-varying signals by temporal discharge patterns or mean firing rates. In particular, we found that a subpopulation of neurons can encode rapidly occurring sounds, such as a click train, with discharges that are not synchronized to individual stimulus events, suggesting a temporal-to-rate transformation. We investigated whether there were stimulus-specific temporal patterns embedded in these seemingly random spike times. Furthermore, we quantitatively analyzed the precision of spike timing at stimulus onset and during ongoing acoustic stimulation. The main findings are the following. 1) Temporal and rate codes may operate at separate stimulus domains or encode the same stimulus domain in parallel via different neuronal populations. 2) Spike timing was crucial to encode stimulus periodicity in "synchronized" neurons. 3) "Nonsynchronized" neurons showed little stimulus-specific spike timing information in their responses to time-varying signals. Such responses therefore represent processed (instead of preserved) information in the auditory cortex. And 4) spike timing on the occurrence of acoustic events was more precise at the first event than at successive ones and more precise with sparsely distributed events (longer time intervals between events) than with densely packed events. These results indicate that auditory cortical neurons mark sparse acoustic events (or onsets) with precise spike timing and transform rapidly occurring acoustic events into firing rate-based representations.

Gabor analysis of auditory midbrain receptive fields: spectro-temporal and binaural composition

The spectro-temporal receptive field (STRF) is a model representation of the excitatory and inhibitory integration area of auditory neurons. Recently it has been used to study spectral and temporal aspects of monaural integration in auditory centers. Here we report the properties of monaural STRFs and the relationship between ipsi- and contralateral inputs to neurons of the central nucleus of cat inferior colliculus (ICC) of cats. First, we use an optimal singular-value decomposition method to approximate auditory STRFs as a sum of time-frequency separable Gabor functions. This procedure extracts nine physiologically meaningful parameters. The STRFs of approximately 60% of collicular neurons are well described by a time-frequency separable Gabor STRF model, whereas the remaining neurons exhibited obliquely oriented or multiple excitatory/inhibitory subfields that require a nonseparable Gabor fitting procedure. Parametric analysis reveals distinct spectro-temporal tradeoffs in receptive field size and modulation filtering resolution. Comparisons between an identical model used to study spatio-temporal integration areas of visual neurons further shows that auditory and visual STRFs share numerous structural properties. We then use the Gabor STRF model to compare quantitatively receptive field properties of contra- and ipsilateral inputs to the ICC. We show that most interaural STRF parameters are highly correlated bilaterally. However, the spectral and temporal phases of ipsi- and contralateral STRFs often differ significantly. This suggests that activity originating from each ear share various spectro-temporal response properties such as their temporal delay, bandwidth, and center frequency but have shifted or interleaved patterns of excitation and inhibition. These differences in converging monaural receptive fields expand binaural processing capacity beyond interaural time and intensity aspects and may enable colliculus neurons to detect disparities in the spectro-temporal composition of the binaural input.

A unifying basis of auditory thresholds based on temporal summation

Thresholds of auditory-nerve (AN) fibers and auditory neurons are commonly specified in terms of sound pressure only, implying that they are independent of time. At the perceptual level, however, the sound pressure required for detection decreases with increasing stimulus duration, suggesting that the auditory system integrates sound over time. The quantity commonly believed to be integrated is sound intensity, implying that the auditory system would have an energy threshold. However, leaky integrators of intensity with time constants of hundreds of milliseconds are required to fit the data. Such time constants are unknown in physiology and are also incompatible with the high temporal resolution of the auditory system, creating the resolution-integration paradox. Here we demonstrate that cortical and perceptual responses are based on integration of the pressure envelope of the sound, as we have previously shown for AN fibers, rather than on intensity. The functions relating the pressure envelope integration thresholds and time for AN fibers, cortical neurons, and perception in the same species (cat), as well as for perception in many different vertebrate species, are remarkably similar. They are well described by a power law that resolves the resolution-integration paradox. The data argue for the integrator to be located in the first synapse in the auditory pathway and we discuss its mode of operation.

Highly variable spike trains underlie reproducible sensorimotor responses in the medicinal leech

The nervous system of the leech is a particularly suitable model to investigate neural coding of sensorimotor responses because it allows both observation of behavior and the simultaneous measurement of a large fraction of its underlying neuronal activity. In this study, we used a combination of multielectrode recordings, videomicroscopy, and computation of the optical flow to investigate the reproducibility of the motor response caused by local mechanical stimulation of the leech skin. We analyzed variability at different levels of processing: mechanosensory neurons, motoneurons, muscle activation, and behavior. Spike trains in mechanosensory neurons were very reproducible, unlike those in motoneurons. The motor response, however, was reproducible because of two distinct biophysical mechanisms. First, leech muscles contract slowly and therefore are poorly sensitive to the jitter of motoneuron spikes. Second, the motor response results from the coactivation of a population of motoneurons firing in a statistically independent way, which reduces the variability of the population firing. These data show that reproducible spike trains are not required to sustain reproducible behaviors and illustrate how the nervous system can cope with unreliable components to produce reliable action.

A unified mechanism for spontaneous-rate and first-spike timing in the auditory nerve

Recent physiological experiments have provided detailed descriptions of the properties of first-spike latency and variability in auditory cortex and nerve in response to pure tones with different envelopes. The envelope-dependence of first-spike timing and precision in auditory cortical neurons appears to reflect properties established in the nerve. First-spike latency properties in individual auditory nerve fibers are strongly correlated with their spontaneous rate (SR). It is shown here that a minimal, plausible model of auditory transduction with two free parameters accurately reproduces the physiological data from the auditory nerve population. The model consists of a simple gain stage, a bandpass filter, a rectifying saturating non-linearity, and a lowpass filter in series. The output of the lowpass filter drives an inhomogeneous Poisson process. The shape of the non-linearity is determined by SR; in physiological terms, this shape depends upon the resting sensitivity of the synapse between the inner hair cell and the auditory nerve. An alternative model for SR generation, where SR is added to the stimulus-driven output of a fixed nonlinearity, fails to account for the data. The results provide a novel, comprehensive and physiologically-based explanation for the range of experimental results on the envelope-dependence of first-spike latency and precision, and its relationship with SR, in the auditory system.

Effects of noise and cue enhancement on neural responses to speech in auditory midbrain, thalamus and cortex

Speech perception depends on the auditory system's ability to extract relevant acoustic features from competing background noise. Despite widespread acknowledgement that noise exacerbates this process, little is known about the neurophysiologic mechanisms underlying the encoding of speech in noise. Moreover, the relative contribution of different brain nuclei to these processes has not been fully established. To address these issues, aggregate neural responses were recorded from within the inferior colliculus, medial geniculate body and over primary auditory cortex of anesthetized guinea pigs to a synthetic vowel-consonant-vowel syllable /ada/ in quiet and in noise. In noise the onset response to the stop consonant /d/ was reduced or eliminated at each level, to the greatest degree in primary auditory cortex. Acoustic cue enhancements characteristic of 'clear' speech (lengthening the stop gap duration and increasing the intensity of the release burst) improved the neurophysiologic representation of the consonant at each level, especially at the cortex. Finally, the neural encoding of the vowel segment was evident at subcortical levels only, and was more resistant to noise than encoding of the dynamic portion of the consonant (release burst and formant transition). This experiment sheds light on which speech-sound elements are poorly represented in noise and demonstrates how acoustic modifications to the speech signal can improve neural responses in a normal auditory system. Implications for understanding neurophysiologic auditory signal processing in children with perceptual impairments and the design of efficient perceptual training strategies are also discussed.

Precision of neural timing: effects of convergence and time-windowing

We study the improvement in timing accuracy in a neural system having n identical input neurons projecting to one target neuron. The n input neurons receive the same stimulus but fire at stochastic times selected from one of four specified probability densities, f, each with standard deviation 1.0 msec. The target cell fires if and when it receives m inputs within a time window of epsilon msec. Let sigma(n,m,epsilon) denote the standard deviation of the time of firing of the target neuron (i.e. the standard deviation of the target neuron's latency relative to the arrival time of the stimulus). Mathematical analysis shows that sigma(n,m,epsilon) is a very complicated function of n, m, and epsilon. Typically, sigma(n,m,epsilon) is a non-monotone function of m and epsilon and the improvement of timing accuracy is highly dependent of the shape of the probability density for the time of firing of the input neurons. For appropriate choices of m, epsilon, and f, the standard deviation sigma(n,m,epsilon) may be as low as 1/n. Thus, depending on these variables, remarkable improvements in timing accuracy of such a stochastic system may occur.

Temporal and spatial coding of periodicity information in the inferior colliculus of awake chinchilla (Chinchilla laniger)

Amplitude modulation responses and onset latencies of multi-unit recordings and evoked potentials were investigated in the central nucleus of inferior colliculus (ICC) in the awake chinchilla. Nine hundred and one recording sites with best frequencies between 60 and 30 kHz showed either phasic (18%), tonic (25%), or phasic-tonic (57%) responses. Of 554 sites tested for responses to modulation frequencies 73% were responsive and 57% showed clear preference for a narrow range of modulation frequencies. Well defined bandpass characteristics were found for 32% of rate modulation transfer functions (rate-MTFs) and 36% of synchronization MTFs (sync-MTFs). The highest best modulation frequency (BMF) of a bandpass rate-MTF was 600 Hz. Neurons with phasic responses to best-frequency tones showed strong phase coupling to modulation frequencies and were dominated by bandpass rate-MTFs and sync-MTFs. Most neurons with tonic responses showed bandpass tuning only for sync-MTFs. Both BMFs and onset latencies changed systematically across frequency-band laminae of the ICC. Low BMFs and long latencies were located medially and high BMFs and short latencies laterally. Latency distributions obtained with evoked potentials to clicks showed a similar gradient to the multi-unit data. These findings are in line with previous findings in different animals including humans and support the hypothesis that temporal processing results in a topographic arrangement orthogonal to the spectral processing axis, thus forming a second neural axis of the auditory system.

Neural representations of sinusoidal amplitude and frequency modulations in the primary auditory cortex of awake primates

We investigated neural coding of sinusoidally modulated tones (sAM and sFM) in the primary auditory cortex (A1) of awake marmoset monkeys, demonstrating that there are systematic cortical representations of embedded temporal features that are based on both average discharge rate and stimulus-synchronized discharge patterns. The rate-representation appears to be coded alongside the stimulus-synchronized discharges, such that the auditory cortex has access to both rate and temporal representations of the stimulus at high and low frequencies, respectively. Furthermore, we showed that individual auditory cortical neurons, as well as populations of neurons, have common features in their responses to both sAM and sFM stimuli. These results may explain the similarities in the perception of sAM and sFM stimuli as well as the different perceptual qualities effected by different modulation frequencies. The main findings include the following. 1) Responses of cortical neurons to sAM and sFM stimuli in awake marmosets were generally much stronger than responses to unmodulated tones. Some neurons responded to sAM or sFM stimuli but not to pure tones. 2) The discharge rate-based modulation transfer function typically had a band-pass shape and was centered at a preferred modulation frequency (rBMF). Population-averaged mean firing rate peaked at 16- to 32-Hz modulation frequency, indicating that the A1 was maximally excited by this frequency range of temporal modulations. 3) Only approximately 60% of recorded units showed statistically significant discharge synchrony to the modulation waveform of sAM or sFM stimuli. The discharge synchrony-based best modulation frequency (tBMF) was typically lower than the rBMF measured from the same neuron. The distribution of rBMF over the population of neurons was approximately one octave higher than the distribution of tBMF. 4) There was a high degree of similarity between cortical responses to sAM and sFM stimuli that was reflected in both discharge rate- or synchrony-based response measures. 5) Inhibition appeared to be a contributing factor in limiting responses at modulation frequencies above the rBMF of a neuron. And 6) neurons with shorter response latencies tended to have higher tBMF and maximum discharge synchrony frequency than those with longer response latencies. rBMF was not significantly correlated with the minimum response latency.

Central auditory onset responses, and temporal asymmetries in auditory perception

Historically, central auditory responses have been studied for their sensitivity to various parameters of tone and noise burst stimulation, with response rate plotted as a function of the stimulus variable. The responses themselves are often quite brief, and locked in time to stimulus onset. In the stimulus amplitude domain, it has recently become clear that these responses are actually driven by properties of the stimulus' onset transient, and this has had important implications for how we interpret responses to manipulations of tone (or noise) burst plateau level. That finding was important in its own right, but a more general scrutiny of the available neurophysiological and psychophysical evidence reveals that there is a significant asymmetry in the neurophysiological and perceptual processing of stimulus onsets and offsets: sound onsets have a more elaborate neurophysiological representation, and receive a greater perceptual weighting. Hypotheses about origins of the asymmetries, derived independently from psychophysics and from neurophysiology, have in common a response threshold mechanism which adaptively tracks the ongoing level of stimulation.

Auditory steady-state responses to exponential modulation envelopes

OBJECTIVE

This study examined the steady-state responses evoked by tones modulated with exponential envelopes. The hypothesis was that stimuli with envelopes containing more rapid changes would evoke larger responses.

DESIGN

Multiple auditory steady-state responses were recorded simultaneously to eight tonal stimuli, four in each ear. The carrier frequencies of the stimuli ranged from 500 to 6000 Hz and the modulation rates were between 75 and 95 Hz. The modulation envelopes were based on functions using sin' where N was 1, 2, 3, or 4. Setting N to 1 produced the traditional sinusoidal modulation.

RESULTS

Exponential envelopes with N greater than 1 produced larger steady-state responses than a sinusoidal envelope. For amplitude-modulation (AM), exponential envelopes increased response amplitudes by 21% at 55 dB pSPL, and by 29% at 35 dB pSPL. The increases were smaller for carrier frequencies of 1500 to 2000 Hz than for lower and higher carrier frequencies. Latencies calculated from phase data increased significantly with increasing N. This was likely caused by the point of maximal envelope-slope shifting later in time as N increased. For frequency modulation (FM), the steady-state responses did not significantly change with changes in the power of the exponential envelopes.

CONCLUSIONS

When tones are amplitude-modulated with exponential envelopes based on sin(N), the amplitude and latency of the steady-state response increased significantly with increasing N. Using exponential envelopes with N greater than 1 should considerably shorten the time needed for responses to become significant when using steady-state responses in objective audiometry.

Mechanisms underlying the sensitivity of neurons in the lateral superior olive to interaural intensity differences

The initial processing of interaural intensity differences (IIDs), the major cue to the azimuthal location of high-frequency sounds in mammals, is carried out by neurons in the lateral superior olivary nucleus (LSO) that receive excitatory input from the ipsilateral ear and inhibitory input from the contralateral ear (IE neurons). The "latency" hypothesis asserts that it is the effects of intensity differences on the latency, and hence the relative timing, of the synaptic inputs to these neurons that is the basis of their sensitivity to IIDs, while other models assign the major role to changes in the relative amplitude of the inputs. To test the latency hypothesis and to determine the contributions of changes in the relative timing and amplitude of synaptic inputs to the IID sensitivity of LSO neurons, a method was developed of generating sets of stimuli that produced either the same changes in the relative timing of inputs without any change in their amplitude (equivalent interaural time difference stimuli) or the same differences in amplitude without any difference in timing (delay-cancelled IID stimuli) as a given range of IIDs. Data were obtained from a sample of IE neurons in the LSO of anesthetized rats using these stimulus paradigms and click and tone-burst stimuli. For click stimuli, the IID sensitivity of a small proportion of neurons was explained entirely by sensitivity to differences in input timing, but the sensitivity of most neurons reflected either sensitivity to the relative amplitude of inputs or to the joint operation of both factors. In neurons whose sensitivity was tested at a number of different absolute sound pressure levels (SPLs), the relative contributions of the two factors tended to differ at different SPLs. The IID sensitivity of onset responses to tone stimuli could be classified into the same three categories but was explained for a larger proportion of neurons by sensitivity to differences in input timing. The IID sensitivity of the late response component of neurons with sustained responses to tones in all cases reflected sensitivity to the relative amplitude of the inputs. The results confirm the contribution of changes in latency produced by intensity changes to the IID sensitivity of the onset responses of many IE neurons in LSO but require rejection of the strong form of the latency hypothesis, which asserts that this factor alone accounts for such sensitivity.

Temporal integration of sound pressure determines thresholds of auditory-nerve fibers

Current propositions of the quantity of sound driving the central auditory system, specifically around threshold, are diverse and at variance with one another. They include sound pressure, sound power, or intensity, which are proportional to the square of pressure, and energy, i.e., the integral of sound power over time. Here we show that the relevant sound quantity and the nature of the threshold can be obtained from the timing of the first spike of auditory-nerve (AN) fibers after the onset of a stimulus. We reason that the first spike is triggered when the stimulus reaches threshold and occurs with fixed delay thereafter. By probing cat AN fibers with characteristic frequency tones of different sound pressure levels and rise times, we show that the differences in relative timing of the first spike (including latencies >100 msec of fibers with low spontaneous rates) can be well accounted for by essentially linear integration of pressure over time. The inclusion of a constant pressure loss or gain to the integrator improves the fit of the model and also accounts for most of the variation of spontaneous rates across fibers. In addition, there are tight correlations among delay, threshold, and spontaneous rate. First-spike timing cannot be explained by models based on a fixed pressure threshold, a fixed power or intensity threshold, or an energy threshold. This suggests that AN fiber thresholds are best measured in units of pressure by time. Possible mechanisms of pressure integration by the inner hair cell-AN fiber complex are discussed.

Between sound and perception: reviewing the search for a neural code

This review investigates the roles of representation, transformation and coding as part of a hierarchical process between sound and perception. This is followed by a survey of how speech sounds and elements thereof are represented in the activity patterns along the auditory pathway. Then the evidence for a place representation of texture features of sound, comprising frequency, periodicity pitch, harmonicity in vowels, and direction and speed of frequency modulation, and for a temporal and synchrony representation of sound contours, comprising onsets, offsets, voice onset time, and low rate amplitude modulation, in auditory cortex is reviewed. Contours mark changes and transitions in sound and auditory cortex appears particularly sensitive to these dynamic aspects of sound. Texture determines which neurons, both cortical and subcortical, are activated by the sound whereas the contours modulate the activity of those neurons. Because contours are temporally represented in the majority of neurons activated by the texture aspects of sound, each of these neurons is part of an ensemble formed by the combination of contour and texture sensitivity. A multiplexed coding of complex sound is proposed whereby the contours set up widespread synchrony across those neurons in all auditory cortical areas that are activated by the texture of sound.

Auditory edge detection: a neural model for physiological and psychoacoustical responses to amplitude transients

Primary segmentation of visual scenes is based on spatiotemporal edges that are presumably detected by neurons throughout the visual system. In contrast, the way in which the auditory system decomposes complex auditory scenes is substantially less clear. There is diverse physiological and psychophysical evidence for the sensitivity of the auditory system to amplitude transients, which can be considered as a partial analogue to visual spatiotemporal edges. However, there is currently no theoretical framework in which these phenomena can be associated or related to the perceptual task of auditory source segregation. We propose a neural model for an auditory temporal edge detector, whose underlying principles are similar to classical visual edge detector models. Our main result is that this model reproduces published physiological responses to amplitude transients collected at multiple levels of the auditory pathways using a variety of experimental procedures. Moreover, the model successfully predicts physiological responses to a new set of amplitude transients, collected in cat primary auditory cortex and medial geniculate body. Additionally, the model reproduces several published psychoacoustical responses to amplitude transients as well as the psychoacoustical data for amplitude edge detection reported here for the first time. These results support the hypothesis that the response of auditory neurons to amplitude transients is the correlate of psychoacoustical edge detection.

Sensitivity of unanesthetized chinchilla auditory system to noise burst onset, and the effects of carboplatin

The gross near-field responses of the auditory nerve and inferior colliculus to noise burst stimuli were recorded through intracranially implanted electrodes in six unanesthetized chinchillas. Responses were studied as a function of stimulus plateau amplitude and rise time, both before and after a systemic dose of 75 mg/kg of carboplatin. Both recording sites showed sensitivity to stimulus level and rise time. Increases in stimulus level and decreases in stimulus rise time each produced increases in the response magnitude, and decreases in response latency. When the stimuli were re-specified as rate of pressure change at sound onset (Pa/s), the amplitude and latency of responses at each site were found to be a direct function of rate of sound pressure change. These data provide the first confirmation in unanesthetized animals of previous single unit observations in barbiturate-anesthetized cats. Carboplatin treatment resulted in a 20-80% loss of inner hair cells, a modest threshold elevation, and a 50-75% reduction in peak response amplitudes. The general patterns of sensitivity to stimulus level and rise time were not markedly affected by carboplatin, nor was the fashion in which response parameters (amplitude and latency) were ruled by rate of pressure change at sound onset.

Functional organization of squirrel monkey primary auditory cortex: responses to pure tones

The spatial organization of response parameters in squirrel monkey primary auditory cortex (AI) accessible on the temporal gyrus was determined with the excitatory receptive field to pure tone stimuli. Dense, microelectrode mapping of the temporal gyrus in four animals revealed that characteristic frequency (CF) had a smooth, monotonic gradient that systematically changed from lower values (0.5 kHz) in the caudoventral quadrant to higher values (5--6 kHz) in the rostrodorsal quadrant. The extent of AI on the temporal gyrus was approximately 4 mm in the rostrocaudal axis and 2--3 mm in the dorsoventral axis. The entire length of isofrequency contours below 6 kHz was accessible for study. Several independent, spatially organized functional response parameters were demonstrated for the squirrel monkey AI. Latency, the asymptotic minimum arrival time for spikes with increasing sound pressure levels at CF, was topographically organized as a monotonic gradient across AI nearly orthogonal to the CF gradient. Rostral AI had longer latencies (range = 4 ms). Threshold and bandwidth co-varied with the CF. Factoring out the contribution of the CF on threshold variance, residual threshold showed a monotonic gradient across AI that had higher values (range = 10 dB) caudally. The orientation of the threshold gradient was significantly different from the CF gradient. CF-corrected bandwidth, residual Q10, was spatially organized in local patches of coherent values whose loci were specific for each monkey. These data support the existence of multiple, overlying receptive field gradients within AI and form the basis to develop a conceptual framework to understand simple and complex sound coding in mammals.

Parallels between timing of onset responses of single neurons in cat and of evoked magnetic fields in human auditory cortex

Sound onsets constitute particularly salient transients and evoke strong responses from neurons of the auditory system, but in the past, such onset responses have often been analyzed with respect to steady-state features of sounds, like the sound pressure level. Recent electrophysiological studies of single neurons from the auditory cortex of anesthetized cats have revealed that the timing and strength of onset responses are shaped by dynamic stimulus properties at their very onsets. Here we demonstrate with magnetoencephalography that stimulus-response relationships very similar to those of the single neurons are observed in two onset components, N100m and P50m, of auditory evoked magnetic fields (AEFs) from the auditory cortex of awake humans. In response to tones shaped with cosine-squared rise functions, N100m and P50m peak latencies vary systematically with tone level and rise time but form a rather invariant function of the acceleration of the envelope at tone onset. Hence N100m and P50m peak latencies, as well as peak amplitudes, are determined by dynamic properties of the stimuli within the first few milliseconds, though not necessarily by acceleration. The changes of N100m and P50m peak latencies with rise time and level are incompatible with a fixed-amplitude threshold model. The direct comparison of the neuromagnetic and single-neuron data shows that, on average, the variance of the neuromagnetic data is larger by one to two orders of magnitude but that favorable measurements can yield variances as low as those derived from neurons with mediocre precision of response timing. The striking parallels between the response timing of single cortical neurons and of AEFs provides a stronger link between single neuron and population activity.

Latency as a function of intensity in auditory neurons: influences of central processing

The response latencies of sensory neurons typically shorten with increases in stimulus intensity. In the central auditory system this phenomenon should have a significant impact on a number of auditory functions that depend critically on an integration of precisely timed neural inputs. Evidence from previous studies suggests that the auditory system not only copes with the potential problems associated with intensity-dependent latency change, but that it also modifies latency change to shape the response properties of many cells for specific functions. This observation suggests that intensity-dependent latency change may undergo functional transformations along the auditory neuraxis. The goal of our study was to explore these transformations by making a direct, quantitative comparison of intensity-dependent latency change among a number of auditory centers from the lower brainstem to the thalamus. We found two main ways in which intensity-dependent latency change transformed along the neuraxis: (1) the range of latency change increased substantially and (2) one particular type of latency change, which has been suggested to be associated with sensitivity to temporally segregated stimulus components, occurred only at the highest centers tested, the midbrain and thalamus. Additional testing in the midbrain (inferior colliculus) indicated that inhibitory inputs are involved in shaping latency change. Our findings demonstrate that the central auditory system modifies intensity-dependent latency changes. We suggest that these changes may be functionally incorporated, actively enhanced, or modified to suit specific functions of the auditory system.

Temporal discharge patterns evoked by rapid sequences of wide- and narrowband clicks in the primary auditory cortex of cat

The present study investigated neural responses to rapid, repetitive stimuli in the primary auditory cortex (A1) of cats. We focused on two important issues regarding cortical coding of sequences of stimuli: temporal discharge patterns of A1 neurons as a function of inter-stimulus interval and cortical mechanisms for representing successive stimulus events separated by very short intervals. These issues were studied using wide- and narrowband click trains with inter-click intervals (ICIs) ranging from 3 to 100 ms as a class of representative sequential stimuli. The main findings of this study are 1) A1 units displayed, in response to click train stimuli, three distinct temporal discharge patterns that we classify as regions I, II, and III. At long ICIs nearly all A1 units exhibited typical stimulus-synchronized response patterns (region I) consistent with previously reported observations. At intermediate ICIs, no clear temporal structures were visible in the responses of most A1 units (region II). At short ICIs, temporal discharge patterns are characterized by the presence of either intrinsic oscillations (at approximately 10 Hz) or a change in discharge rate that was a monotonically decreasing function of ICI (region III). In some A1 units, temporal discharge patterns corresponding to region III were absent. 2) The boundary between regions I and II (synchronization boundary) had a median value of 39.8 ms ICI ([25%, 75%] = [20.4, 58. 8] ms ICI; n = 131). The median boundary between regions II and III was estimated at 6.3 ms ([25%, 75%] = [5.2, 9.7] ms ICI; n = 47) for units showing rate changes (rate-change boundary). 3) The boundary values between different regions appeared to be relatively independent of stimulus intensity (at modest sound levels) or the bandwidth of the clicks used. 4) There is a weak correlation between a unit's synchronization boundary and its response latency. Units with shorter latencies appeared to also have smaller boundary values. And 5) based on these findings, we proposed a two-stage model for A1 neurons to represent a wide range of ICIs. In this model, A1 uses a temporal code for explicitly representing long ICIs and a rate code for implicitly representing short ICIs.

Discharge patterns of neurons in the ventral nucleus of the lateral lemniscus of the unanesthetized rabbit

The ventral nucleus of the lateral lemniscus (VNLL) is a major auditory nucleus that sends a large projection to the inferior colliculus. Despite its prominence, the responses of neurons in the VNLL have not been extensively studied. Previous studies in nonecholocating species have used anesthesia, which is known to affect discharge patterns. In addition, there is disagreement about the proportion of neurons that are sensitive to binaural stimulation. This report examines the responses of neurons in the VNLL of the unanesthetized rabbit to monaural and binaural stimuli. Most neurons responded to contralateral tone bursts at their best frequency and had either sustained or phasic discharge patterns. A few neurons were only inhibited. Most sustained neurons were classified as short-latency sustained (SL-sustained), but a few were of long latency. Some SL-sustained neurons exhibited multiple peaks in their discharge pattern, i.e., they had a "chopper" discharge pattern, whereas other SL-sustained neurons did not exhibit this pattern. In ordinary chopper neurons, the multiple peaks corresponded to the evenly spaced action potentials of a regular discharge. In unusual chopper neurons, the action potential associated with a particular peak could fail to occur during any one presentation of the stimulus. Unusual chopper neurons had a relatively irregular discharge. Phasic neurons were of two types: onset and transient. Onset neurons typically responded with a single action potential at the onset of the tone, whereas transient neurons produced a burst of action potentials. Transient neurons were relatively rare. About half the neurons also were influenced by ipsilateral stimulation. Most binaurally influenced neurons were either sensitive to interaural temporal disparities (ITDs) or excited by contralateral stimulation and inhibited by ipsilateral stimulation. Neurons sensitive to ITDs were mostly of the onset type and were embedded in the fiber tract medial to the main part of the nucleus. Neurons inhibited by ipsilateral stimulation could be of the sustained or onset type. The sustained neurons were located on the periphery of the main nucleus as well as in the fiber tract. Most of the monaural neurons were in the main, high-density part of VNLL. The present results demonstrate that the VNLL contains neurons with a heterogeneous set of responses, and that many of the neurons are binaural.

Response magnitude and timing of auditory response initiation in the inferior colliculus of the awake chinchilla

Recent single-unit studies in anesthetized cats have revealed that the latency and strength of transient responses to tone burst stimuli are determined largely by stimulus events in the first few ms of the signal. The present study sought to extend these findings by studying the inferior colliculus potential (ICP) in unanesthetized chinchillas. The ICP magnitude and latency were studied as a function of the plateau amplitude and rise time of noise burst stimuli. ICP amplitude increased with stimulus amplitude and decreased with stimulus rise time. ICP latency decreased with stimulus amplitude and increased with stimulus rise time. The absolute values of the ICP latencies confirmed that it is only the first few ms of the stimulus which determine the timing of response initiation, and therefore, that it is not the plateau level of the stimulus that directly determines the latent period. These data constitute a direct link between earlier single-unit studies in anesthetized animals and brainstem-evoked potential data in animals and man.

Further observations on the threshold model of latency for auditory neurons

This paper examines the recent claim of Phillips, that the threshold model of latency is inadequate to account for the changes of latency of auditory cortical neurons in response to tones of different amplitudes and rise times. I argue that Phillips' analysis was based on an incorrect assumption and that he therefore rejected the model for the wrong reasons, though correctly, as the model is in fact inadequate, as demonstrated here and previously. The failure of the model has significant implications for signal processing in the auditory system.

Neuronal coding of interaural transient envelope disparities

Onsets are salient and important transient (i.e. dynamic) features of acoustic signals, and evoke vigorous responses from most auditory neurons, but paradoxically these onset responses have most often been analysed with respect to steady-state stimulus features, e.g. the sound pressure level (SPL). In nearly all studies concerned with the coding of differences in SPL at the two ears (interaural level differences; ILDs), which provide a major cue for the azimuthal location of high frequency sound sources, interaural onset disparities were covaried with ILD, but the possibly confounding effects of this covariation on neuronal responses have been entirely neglected. Therefore, dichotic stimulus paradigms were designed here in which onset and steady-state features were varied independently. Responses were recorded from single neurons in the inferior colliculus of rats, anaesthetized with pentobarbital and xylazine. It is demonstrated that onset responses, or the onset response components of neurons with more complex temporal response patterns, are dependent on the binaural combination of dynamic envelope features associated with conventional ILD stimulus paradigms, but not on the binaural combination of steady-state SPLs reached after the onset. In contrast, late or sustained response components appear more sensitive to the binaural combination of steady-state SPLs. These data stress the general necessity for a separate analysis of onset and late response components, with respect to different stimulus features, and suggest a need for re-evaluation of existing studies on ILD coding. The sensitivity of onset responses to the binaural combination of envelope transients, rather than to steady-state ILD, is in line with their sensitivity to other interaural envelope disparities, created by stationary or moving sounds.

Functional specialization in auditory cortex: responses to frequency-modulated stimuli in the cat's posterior auditory field

The mammalian auditory cortex contains multiple fields but their functional role is poorly understood. Here we examine the responses of single neurons in the posterior auditory field (P) of barbiturate- and ketamine-anesthetized cats to frequency-modulated (FM) sweeps. FM sweeps traversed the excitatory response area of the neuron under study, and FM direction and the linear rate of change of frequency (RCF) were varied systematically. In some neurons, sweeps of different sound pressure levels (SPLs) also were tested. The response magnitude (number of spikes corrected for spontaneous activity) of nearly all field P neurons varied with RCF. RCF response functions displayed a variety of shapes, but most functions were of low-pass characteristic or peaked at rather low RCFs (<100 kHz/s). Neurons with strong responses to high RCFs (high-pass or nonselective RCF response function characteristics) all displayed spike count-SPL functions to tone burst onsets that were monotonic or weakly nonmonotonic. RCF response functions and best RCFs often changed with SPL. For most neurons, FM directional sensitivity, quantified by a directional sensitivity (DS) index, also varied with RCF and SPL, but the mean and width of the distribution of DS indices across all neurons was independent of RCF. Analysis of response timing revealed that the phasic response of a neuron is triggered when the instantaneous frequency of the sweep reaches a particular value, the effective Fi. For a given neuron, values of effective Fi were independent of RCF, but depended on FM direction and SPL and were associated closely with the boundaries of the neuron's frequency versus amplitude response area. The standard deviation (SD) of the latency of the first spike of the response decreased with RCF. When SD was expressed relative to the rate of change of stimulus frequency, the resulting index of frequency jitter increased with RCF and did so rather uniformly in all neurons and largely independent of SPL. These properties suggest that many FM parameters are represented by, and may be encoded in, orderly temporal patterns across different neurons in addition to the strength of responses. When compared with neurons in primary and anterior auditory fields, field P neurons respond better to relatively slow FMs. Together with previous studies of responses to modulations of amplitude, such as tone onsets, our findings suggest more generally that field P may be best suited for processing signals that vary relatively slowly over time.