Some possible neural correlates of combination tones

[The Erlangen model of tinnitus development-New perspective and treatment strategy]

BACKGROUND

About one sixth of the population of western industrialized nations suffers from chronic, subjective tinnitus, causing socioeconomic treatment and follow-up costs of almost 22 billion euros per year in Germany alone. According to the prevailing view, tinnitus develops as a consequence of a maladaptive neurophysiological process in the brain triggered by hearing loss.

OBJECTIVES

The Erlangen model of tinnitus development presented here is intended to propose a comprehensive neurophysiological explanation for the initial occurrence of the phantom sound after hearing loss. Based on the model, a new treatment strategy will be developed.

MATERIALS AND METHODS

The model summarized here is based on various animal and human physiological studies conducted in recent years.

RESULTS

The Erlangen model considers subjective tinnitus as a side effect of a physiological mechanism that permanently optimizes information transmission into the auditory system by means of stochastic resonance (SR) even in the healthy auditory system. In fact, hearing-impaired patients with tinnitus hear better on average than those without tinnitus. This unfamiliar perspective on the phantom percept may already help affected patients to cope better with their suffering. In addition, based on the model, low intensity noise tinnitus suppression (LINTS) has been developed as a new, individually adapted treatment strategy for tonal tinnitus and has already been successfully tested in patients.

CONCLUSIONS

A possible limiting factor for the model and treatment strategy is the pitch of the tinnitus percept, which may require adjustments to the treatment strategy for frequencies above about 5 kHz.

Simultaneous spike-time locking to multiple frequencies

Locking of neural firing is ubiquitously observed in the brain and occurs when neurons fire at a particular phase or in synchronization with an external signal. Here we study in detail the locking of single neurons to multiple distinct frequencies at the example of p-type electroreceptor afferents in the electrosensory system of the weakly electric fish Apteronotus leptorhynchus (brown ghost knifefish). We find that electrosensory afferents and pyramidal cells in the electrosensory lateral line lobe (ELL) lock to multiple frequencies, including the electric organ discharge (EOD) frequency, beat, and stimulus itself. We identify key elements necessary for locking to multiple frequencies, study its limits, and provide concise mathematical models reproducing our main findings. Our findings provide another example of how rate and temporal codes can coexist and complement each other in single neurons and demonstrate that sensory coding in p-type electroreceptor afferents provides a much richer representation of the sensory environment than commonly assumed. Since the underlying mechanisms are not specific to the electrosensory system, our results could provide the basis for studying multiple frequency locking in other systems.NEW & NOTEWORTHY Locking of neuronal spikes to external and internal signals is a ubiquitous neurophysiological mechanism that has been extensively studied in several brain areas and species. Using experimental data from the electrosensory system and concise mathematical models, we analyze how a single neuron can simultaneously lock to multiple frequencies. Our findings demonstrate how temporal and rate codes can complement each other and lead to rich neuronal representations of sensory signals.

Learning Pitch with STDP: A Computational Model of Place and Temporal Pitch Perception Using Spiking Neural Networks

Pitch perception is important for understanding speech prosody, music perception, recognizing tones in tonal languages, and perceiving speech in noisy environments. The two principal pitch perception theories consider the place of maximum neural excitation along the auditory nerve and the temporal pattern of the auditory neurons’ action potentials (spikes) as pitch cues. This paper describes a biophysical mechanism by which fine-structure temporal information can be extracted from the spikes generated at the auditory periphery. Deriving meaningful pitch-related information from spike times requires neural structures specialized in capturing synchronous or correlated activity from amongst neural events. The emergence of such pitch-processing neural mechanisms is described through a computational model of auditory processing. Simulation results show that a correlation-based, unsupervised, spike-based form of Hebbian learning can explain the development of neural structures required for recognizing the pitch of simple and complex tones, with or without the fundamental frequency. The temporal code is robust to variations in the spectral shape of the signal and thus can explain the phenomenon of pitch constancy.

Pitch is the perceptual correlate of sound frequency. Our auditory system has a sophisticated mechanism to process and perceive the neural information corresponding to pitch. This mechanism employs both the place and the temporal pattern of pitch-evoked neural events. Based on the known functions of the auditory system, we develop a computational model of pitch perception using a network of neurons with modifiable connections. We demonstrate that a well-known neural learning rule that is based on the timing of the neural events can identify and strengthen the neuronal connections that are most effective for the extraction of pitch. By providing an insight into how the auditory system interprets pitch information, the results of our study can be used to develop improved sound processing strategies for cochlear implants. In cochlear implant hearing, auditory percept is generated by stimulating the auditory neurons with controlled electrical impulses, enhancing which with the help of the model would lead to a better representation of pitch and would subsequently improve music perception and speech understanding in noisy environments in cochlear implant users.

Implications of within-fiber temporal coding for perceptual studies of F0 discrimination and discrimination of harmonic and inharmonic tone complexes

Recent psychophysical studies suggest that normal-hearing (NH) listeners can use acoustic temporal-fine-structure (TFS) cues for accurately discriminating shifts in the fundamental frequency (F0) of complex tones, or equal shifts in all component frequencies, even when the components are peripherally unresolved. The present study quantified both envelope (ENV) and TFS cues in single auditory-nerve (AN) fiber responses (henceforth referred to as neural ENV and TFS cues) from NH chinchillas in response to harmonic and inharmonic complex tones similar to those used in recent psychophysical studies. The lowest component in the tone complex (i.e., harmonic rank N) was systematically varied from 2 to 20 to produce various resolvability conditions in chinchillas (partially resolved to completely unresolved). Neural responses to different pairs of TEST (F0 or frequency shifted) and standard or reference (REF) stimuli were used to compute shuffled cross-correlograms, from which cross-correlation coefficients representing the degree of similarity between responses were derived separately for TFS and ENV. For a given F0 shift, the dissimilarity (TEST vs. REF) was greater for neural TFS than ENV. However, this difference was stimulus-based; the sensitivities of the neural TFS and ENV metrics were equivalent for equal absolute shifts of their relevant frequencies (center component and F0, respectively). For the F0-discrimination task, both ENV and TFS cues were available and could in principle be used for task performance. However, in contrast to human performance, neural TFS cues quantified with our cross-correlation coefficients were unaffected by phase randomization, suggesting that F0 discrimination for unresolved harmonics does not depend solely on TFS cues. For the frequency-shift (harmonic-versus-inharmonic) discrimination task, neural ENV cues were not available. Neural TFS cues were available and could in principle support performance in this task; however, in contrast to human-listeners' performance, these TFS cues showed no dependence on N. We conclude that while AN-fiber responses contain TFS-related cues, which can in principle be used to discriminate changes in F0 or equal shifts in component frequencies of peripherally unresolved harmonics, performance in these two psychophysical tasks appears to be limited by other factors (e.g., central processing noise).

Do 'Dominant Frequencies' explain the listener's response to formant and spectrum shape variations?

Psychoacoustic experimentation shows that formant frequency shifts can give rise to more significant changes in phonetic vowel timber than differences in overall level, bandwidth, spectral tilt, and formant amplitudes. Carlson and Granström's perceptual and computational findings suggest that, in addition to spectral representations, the human ear uses temporal information on formant periodicities ('Dominant Frequencies') in building vowel timber percepts. The availability of such temporal coding in the cat's auditory nerve fibers has been demonstrated in numerous physiological investigations undertaken during recent decades. In this paper we explore, and provide further support for, the Dominant Frequency hypothesis using KONVERT, a computational auditory model. KONVERT provides auditory excitation patterns for vowels by performing a critical-band analysis. It simulates phase locking in auditory neurons and outputs DF histograms. The modeling supports the assumption that listeners judge phonetic distance among vowels on the basis formant frequency differences as determined primarily by a time-based analysis. However, when instructed to judge psychophysical distance among vowels, they can also use spectral differences such as formant bandwidth, formant amplitudes and spectral tilt. Although there has been considerable debate among psychoacousticians about the functional role of phase locking in monaural hearing, the present research suggests that detailed temporal information may nonetheless play a significant role in speech perception.

Response properties of cochlear nucleus neurons in monkeys

Much of what is known about how the cochlear nuclei participate in mammalian hearing comes from studies of non-primate mammalian species. To determine to what extent the cochlear nuclei of primates resemble those of other mammalian orders, we have recorded responses to sound in three primate species: marmosets, cynomolgus macaques, and squirrel monkeys. These recordings show that the same types of temporal firing patterns are found in primates that have been described in other mammals. Responses to tones of neurons in the ventral cochlear nucleus have similar tuning, latencies, post-stimulus time and interspike interval histograms as those recorded in non-primate cochlear nucleus neurons. In the dorsal cochlear nucleus, too, responses were similar. From these results it is evident that insights gained from non-primate studies can be applied to the peripheral auditory system of primates.

Basilar-membrane response to multicomponent stimuli in chinchilla

The response of chinchilla basilar membrane in the basal region of the cochlea to multicomponent (1, 3, 5, 6, or 7) stimuli was studied using a laser interferometer. Three-component stimuli were amplitude-modulated signals with modulation depths that varied from 25% to 200% and the modulation frequency varied from 100 to 2000 Hz while the carrier frequency was set to the characteristic frequency of the region under study (approximately 6.3 to 9 kHz). Results indicate that, for certain modulation frequencies and depths, there is enhancement of the response. Responses to five equal-amplitude sine wave stimuli indicated the occurrence of nonlinear phenomena such as spectral edge enhancement, present when the frequency spacing was less than 200 Hz, and mutual suppression. For five-component stimuli, the first, third, or fifth component was placed at the characteristic frequency and the component frequency separation was varied over a 2-kHz range. Responses to seven component stimuli were similar to those of five-component stimuli. Six-component stimuli were generated by leaving out the center component of the seven-component stimuli. In the latter case, the center component was restored in the basilar-membrane response as a result of distortion-product generation in the nonlinear cochlea.

Physical basis of two-tone interference in hearing

The cochlea uses active amplification to capture faint sounds. It has been proposed that the amplifier comprises a set of self-tuned critical oscillators: each hair cell contains a force-generating dynamical system that is maintained at the threshold of an oscillatory instability, or Hopf bifurcation. While the active response to a pure tone provides frequency selectivity, exquisite sensitivity, and wide dynamic range, its intrinsic nonlinearity causes tones of different frequency to interfere with one another in the cochlea. Here we determine the response to two tones, which provides a framework for understanding how the ear processes the more complex sounds of speech and music. Our calculations of two-tone suppression and the spectrum of distortion products generated by a critical oscillator accord with experimental observations of basilar membrane motion and the nervous response. We discuss how the response of a set of self-tuned oscillators, covering a range of characteristic frequencies, represents the structure of a complex sound. The frequency components of the stimulus can be inferred from the timing of neural spikes elicited by the vibrating hair cells. Passive prefiltering by the basilar membrane improves pitch discrimination by reducing interference between tones. Our analysis provides a general framework for examining the relation between the physical nature of the peripheral detection apparatus and psychophysical phenomena such as the sensation of dissonance and auditory illusions.

Temporal coding of 200% amplitude modulated signals in the ventral cochlear nucleus of cat

The quasiperiodicity in the acoustic waveform in speech and music is a pervasive feature in our acoustic environment. The use of 200% amplitude modulated (AM) signals allows the study of rate and temporal envelope coding using three equal amplitude components, a situation that is frequently approximated in natural vocalizations. The recordings reported here were made in the ventral cochlear nucleus of the cat, a site of auditory signal feature enhancement and the origin of several ascending auditory pathways. The discharge rate vs modulation frequency relation was nearly always all-pass in shape for all unit types indicating that discharge rate is not a code for modulation frequency. Onset cells, especially onset-choppers and onset-I units, exhibited remarkable phase locking to the signal envelope, nearly to the exclusion of phase locking to the AM components. They exhibited lowpass temporal modulation transfer functions (tMTF) that occasionally had corner frequencies greater than 1 kHz. Primary-like, primary-like with notch, and onset-L units all exhibited considerable variability in their coding properties with tMTFs that varied from lowpass to bandpass in shape. The bandpass shape became more frequent with increasing stimulus levels. A common feature in cochlear nucleus units was less sensitivity to the level of the AM stimulus than is present in the auditory nerve. Phase locking to the envelope persisted over a wider range of stimulus levels than rate changes in a subset of the units studied. The tMTFs for a 100% sinusoidally modulated, spectrally-flat noise was similar in amplitude and bandwidth to those obtained for AM stimuli. The tMTF was relatively insensitive to carrier frequencies different than the unit characteristic frequency. AM synchrony vs level curves exhibited systematic shifts that equaled or exceeded dynamic rate shifts that occur with increasing levels of a noise masker. Phase locking to the envelope was robust under a wide variety of signal conditions in all unit types. The ordering of response types based on the maximum of the tMTF is onset-I = onset-chop > choppers = primarylike-with-notch = onset-L > primarylike.

Stochastic resonance in a single neuron model: theory and analog simulation

Here, we consider a noisy, bistable, single neuron model in the presence of periodic external modulation. The modulation induces a correlated switching between states driven by the noise. The information flow through the system, from the modulation, or signal, to the output switching events, leads to a succession of strong peaks in the power spectrum. The signal-to-noise ratio (SNR) obtained from this power spectrum is a measure of the information content in the neuron response. With increasing noise intensity, the SNR passes through a maximum: an effect which has been called stochastic resonance, and which was first advanced as a possible explanation of the observed periodicity in the recurrences of the Earth's ice ages. We treat the problem within the framework of a recently developed approximate theory, valid in the limits of weak noise intensity, weak periodic forcing and low forcing frequency, for both additive and multiplicative noise. Moreover, we have constructed an analog simulator of the neuron which demonstrates the stochastic resonance effect, and with which we have measured the SNRs for comparison with the theoretical results. Our model should be of interest in situations where a single inherently noisy neuron is the receptor of a periodic signal, which is itself noisy, either from the network or from an external source.

Off-responses in the auditory system in relation to the signal end phase and neuronal characteristic frequency

Off-responses of single neurons from the inferior colliculus (IC) as well as summed off-responses (evoked potentials) from the IC, cochlear nucleus and auditory nerve were studied while varying the end phase of tonal signals. It was found that with tonal signals higher than the characteristic frequency, off-responses at all the auditory levels studied were greatest at the end phases near 0 and 180 degrees and were of minimal value at the end phases near 90 and 270 degrees. On the contrary, with tonal signals lower than the characteristic frequency, the greatest off-responses corresponded to the end phases near 90 and 270 degrees, and responses of the lowest value were registered at the end phases near 0 and 180 degrees. The observed phenomena did not result from transient 'off'-responses of the acoustic system. It is suggested that they reflect certain cochlear processes connected with different effectiveness of oscillation phases at frequencies below and above the resonance.

Wiener kernel analysis of responses from anteroventral cochlear nucleus neurons

Responses to pseudo-random Gaussian white noise, tones and clics were recorded from neurons in the anteroventral cochlear nucleus (AVCN) of barbiturate anesthetized cats. The responses to white noise were used to calculate estimates of the zero-, first- and second-order Wiener kernels for these neurons. The Wiener kernels did contain useful information on the fundamental, DC and second harmonic components of the responses of AVCN neurons to tones, clicks and noise. However, they generally did not provide predictions of the difference tone distortion products found in the peripheral auditory system. Overall, the addition of the second kernel improved a prediction based on the zero- and first-order kernels, but not by very much. If the estimates of the Wiener kernels were not very good, then a second-order prediction could be worse than a first-order one. To produce good estimates of the Wiener kernels, many repetitions of very long Gaussian white noise stimuli are necessary. Therefore the technique does not permit rapid data collection. Further, exposure to long duration high intensity noise can result in acoustic trauma. This damage effects the mechanism that generates the difference tone distortion products, and it can also affect the tuning of the auditory neurons. Thus Wiener's nonlinear system identification theory has only limited usefulness in the analysis of the peripheral auditory system.

Neural coding of repetitive clicks in the medial geniculate body of cat

The activity of 418 medial geniculate body (MGB) units was studied in response to repetitive acoustic pulses in 35 nitrous oxide anaesthetized cats. The proportion of MGB neurons insensitive to repetitive clicks was close to 30%. On the basis of their pattern of discharge, the responsive units were divided into three categories. The majority of them (71%), classified as "lockers', showed discharges precisely time-locked to the individual clicks of the train. A few units (8%), called "groupers', had discharges loosely synchronized to low-rate repetitive clicks. When the spikes were not synchronized, the cell had transient or sustained responses for a limited frequency range and was classified as a "special responder' (21%). Responses of "lockers' were time-locked up to a limiting rate, which varied between 10 and 800 Hz; half of the "lockers' had a limiting rate of locking equal to or higher than 100 Hz. The degree of entrainment, defined as the probability that each click evokes at least one spike, regularly decreases for increasing rates; on the other hand, the precision of locking increasing increases with frequency. The time jitter observed at 100 Hz might be as small as 0.2 ms and was 1.2 ms on average. The population of "lockers' can mark with precision the transients of complex sounds and has response properties still compatible with a temporal coding of the fundamental frequency of most animal vocalizations.

Conditions under which a number of sinusoids may be instantaneously in phase

The literature on the phase relationships between frequency components of a Fourier analysis is reviewed, with examples and theories from acoustics and neurophysiology. Given n sinusoids of different frequencies omega1, omega2, .., omegan and phase angles phi1, phi2, .., phin, it is shown that for n greater than or equal to 2 the set of initial phase angles allowing the n sinusoids to be in phase at some time t0 consists of one or more planes of constant dimension 2 and that for n = 2 such a time t0 always exists. The conditions under which the common phase of n sinusoids at one time t0 will be the same as the common phase at another time t0 are also investigated. The importance of incommensurately related frequency components is emphasized by proofs which do not depend on harmonic relationships. Proofs are formulated in a linear algebra format to demonstrate the versatility of the method for analysing long sequences of frequencies and phases.

Some experiments relating to the perception of complex tones

Pattern recognition models for the perception of complex tones assume that the pitch of a complex tone is derived from more primary sensations, such as the pitches of the individual partials. Thus a complex tone will only have a well-defined pitch when at least one partial in the complex is separately perceptible. Models based on time-interval measurements, on the other hand, require an interaction of the original components, so that the periodicity of the input waveform is preserved. In Experiment I the relative intensity of a “target” tone, necessary for its identification in the presence of either one or two “masking” tones, was determined, over a range of frequencies. This intensity changes abruptly at around 5 kHz, a result consistent with the idea that the pitches of pure tones are determined by temporal mechanisms for frequencies up to 5 kHz, and by place mechanisms for frequencies above this. In Experiments II and III the audibility of the partials in a multi-tone complex was measured as a function of their frequency separation and compared with the range of conditions over which a complex stimulus produced a clear pitch sensation, using the same set of subjects in each experiment. It was found that under some conditions the complex had a well-defined pitch when none of the individual partials was separately audible. This is contrary to the predictions from the pattern recognition models. The effects of masking noise in the frequency region below the complex, and the results of individual subjects, also did not conform with the predictions from these models. Such models are not ruled out, however, for low harmonic numbers, or for stimuli containing only a small number of partials.

Neural encoding of sound sensation evoked by electrical stimulation of the acoustic nerve

A series of psychoacoustic experiments was conducted in subjects implanted with a permanent intracochlear bipolar electrode. These experiments were designed to reveal the nature of the sensation evoked by direct sinusoidal electrical stimulation of the acoustic nerve. A series of single unit experiments in the inferior colliculus of cats was then conducted, using intracochlear stimulus electrodes identical to those implanted in human subjects in all respects except size, and using identical stimuli. These physiological experiments were designed to reveal how sounds evoked by intracochlear electrical stimulation in humans are generated and encoded in the auditory nervous system. Among the results were the following: 1) The sensation arises from direct electrical stimulation of the acoustic nerve. It is not “electrophonic” hearing arising from electro-mechanical excitation of hair cells. 2) While sounds are heard with electrical stimulation at frequencies from below 25 to above 10,000 Hz, the useful range of discriminative hearing is limited to frequencies below 400–600 Hz. 3) There is no “place” coding of electrical stimuli of different frequency. Tonal sensations generated by electrical stimulation must be encoded by the time order of discharge of auditory neurons. 4) The periods of sinusoidal electrical stimuli are encoded in discharges of inferior colliculus neurons at frequencies up to 400–600 Hz. 5) Both psychoacoustic and physiological evidence indicates that the low tone sensations evoked by electrical stimulation are akin to the sensations of “periodicity pitch” generated in the normal cochlea. 6) Most cochlear hair cells are lost with intracochlear implantation with this electrode. Most ganglion cells survive implantation. Implications of these experiments for further development of an acoustic prosthesis are discussed.

Dynamics of encoding in a population of neurons

A simple encoder model, which is a reasonable idealization from known electrophysiological properties, yields a population in which the variation of the firing rate with time is a perfect replica of the shape of the input stimulus. A population of noise-free encoders which depart even slightly from the simple model yield a very much degraded copy of the input stimulus. The presence of noise improves the performance of such a population. The firing rate of a population of neurons is related to the firing rate of a single member in a subtle way.

Coding of sounds in lower levels of the auditory system

The great number of investigations and advanced developments in neurophysiology and psychoacoustics during recent years have extensively increased our knowledge about the frequency analysis of simple sounds in the peripheral auditory system.

New methods have facilitated quantitative measurements of the amplitude of the submicroscopic vibration of a narrow segment of the basilar membrane in anaesthetized animals at physiological sound intensities. The results of these studies have quantitatively confirmed the results of past studies by showing that the basilar membrane has a selectivity with regard to tone frequency. In addition to this, the recent studies have increased our knowledge about the finer details of vibration of the basilar membrane. At the lowest levels used in the recent investigations, i.e. about 70 dB SPL, the selectivity in the 7 kHz region of the basilar membrane was found to be greater than expected on the basis of extrapolation of older data. Moreover, the high frequency slope of the tuning curves of the basilar membrane was found to be particularly steep. The results of these recent studies, furthermore, showed that the basilar membrane vibrates in a non-linear way at intensities within the physiological range. This non-linearity results in a broadening of the selectivity curves of a narrow segment of the basilar membrane when the sound intensity is increased.

Little is known as to how the motion of the basilar membrane is transformed to excitation of the cochlear sensory cells, i.e. the haircells. The excitation may be related to displacement, spatial differentiation or other transformations of the basilar membrane motion. Recording from the interior of mammalian haircells has so far been unsuccessful, and the neural excitatory process within the haircells in the cochlea is as yet practically unknown. Studies of the haircells in the lateral line organ of fish have provided fundamental knowledge about their excitation; since they in many respects resemble those in the mammalian cochlea, the results very probably can be applied to the excitatory process in the mammalian cochlea.