Behavioral and modeling studies of sound localization in cats: effects of stimulus level and duration

Heterogeneous spatial tuning in the auditory pathway of the Mongolian Gerbil (Meriones unguiculatus)

Sound-source localization is based on spatial cues arising due to interactions of sound waves with the torso, head and ears. Here, we evaluated neural responses to free-field sound sources in the central nucleus of the inferior colliculus (CIC), the medial geniculate body (MGB) and the primary auditory cortex (A1) of Mongolian gerbils. Using silicon probes we recorded from anaesthetized gerbils positioned in the centre of a sound-attenuating, anechoic chamber. We measured rate-azimuth functions (RAFs) with broad-band noise of varying levels presented from loudspeakers spanning 210° in azimuth and characterized RAFs by calculating spatial centroids, Equivalent Rectangular Receptive Fields (ERRFs), steepest slope locations and spatial-separation thresholds. To compare neuronal responses with behavioural discrimination thresholds from the literature we performed a neurometric analysis based on signal-detection theory. All structures demonstrated heterogeneous spatial tuning with a clear dominance of contralateral tuning. However, the relative amount of contralateral tuning decreased from the CIC to A1. In all three structures spatial tuning broadened with increasing sound-level. This effect was strongest in CIC and weakest in A1. Neurometric spatial-separation thresholds compared well with behavioural discrimination thresholds for locations directly in front of the animal. Our findings contrast with those reported for another rodent, the rat, which exhibits homogenous and sharply delimited contralateral spatial tuning. Spatial tuning in gerbils resembles more closely the tuning reported in A1 of cats, ferrets and non-human primates. Interestingly, gerbils, in contrast to rats, share good low-frequency hearing with carnivores and non-human primates, which may account for the observed spatial tuning properties.

Population rate-coding predicts correctly that human sound localization depends on sound intensity

Human sound localization is an important computation performed by the brain. Models of sound localization commonly assume that sound lateralization from interaural time differences is level invariant. Here we observe that two prevalent theories of sound localization make opposing predictions. The labelled-line model encodes location through tuned representations of spatial location and predicts that perceived direction is level invariant. In contrast, the hemispheric-difference model encodes location through spike-rate and predicts that perceived direction becomes medially biased at low sound levels. Here, behavioral experiments find that softer sounds are perceived closer to midline than louder sounds, favoring rate-coding models of human sound localization. Analogously, visual depth perception, which is based on interocular disparity, depends on the contrast of the target. The similar results in hearing and vision suggest that the brain may use a canonical computation of location: encoding perceived location through population spike rate relative to baseline.

Being able to localize sounds helps us make sense of the world around us. The brain works out sound direction by comparing the times of when sound reaches the left versus the right ear. This cue is known as interaural time difference, or ITD for short. But how exactly the brain decodes this information is still unknown.

The brain contains nerve cells that each show maximum activity in response to one particular ITD. One idea is that these nerve cells are arranged in the brain like a map from left to right, and that the brain then uses this map to estimate sound direction. This is known as the Jeffress model, after the scientist who first proposed it. There is some evidence that birds and alligators actually use a system like this to localize sounds, but no such map of nerve cells has yet been identified in mammals. An alternative possibility is that the brain compares activity across groups of ITD-sensitive nerve cells. One of the oldest and simplest ways to measure this is to compare nerve activity in the left and right hemispheres of the brain. This readout is known as the hemispheric difference model.

By analyzing data from published studies, Ihlefeld, Alamatsaz, and Shapley discovered that these two models make opposing predictions about the effects of volume. The Jeffress model predicts that the volume of a sound will not affect a person’s ability to localize it. By contrast, the hemispheric difference model predicts that very soft sounds will lead to systematic errors, so that for the same ITD, softer sounds are perceived closer towards the front than louder sounds. To investigate this further, Ihlefeld, Alamatsaz, and Shapley asked healthy volunteers to localize sounds of different volumes. The volunteers tended to mis-localize quieter sounds, believing them to be closer to the body’s midline than they actually were, which is inconsistent with the predictions of the Jeffress model.

These new findings also reveal key parallels to processing in the visual system. Visual areas of the brain estimate how far away an object is by comparing the input that reaches the two eyes. But these estimates are also systematically less accurate for low-contrast stimuli than for high-contrast ones, just as sound localization is less accurate for softer sounds than for louder ones. The idea that the brain uses the same basic strategy to localize both sights and sounds generates a number of predictions for future studies to test.

Audiovisual Interactions in Front and Rear Space

The human visual and auditory systems do not encode an entirely overlapped space when static head and body position are maintained. While visual capture of sound source location in the frontal field is known to be immediate and direct, visual influence in the rear auditory space behind the subject remains under-studied. In this study we investigated the influence of presenting frontal LED flashes on the perceived location of a phantom sound source generated using time-delay-based stereophony. Our results show that frontal visual stimuli affected auditory localization in two different ways - (1) auditory responses were laterally shifted (left or right) toward the location of the light stimulus and (2) auditory responses were more often in the frontal field. The observed visual effects do not adhere to the spatial rule of multisensory interaction with regard to the physical proximity of cues. Instead, the influence of visual cues interacted closely with front-back confusions in auditory localization. In particular, visually induced shift along the left-right direction occurred most often when an auditory stimulus was localized in the same (frontal) field as the light stimulus, even when the actual sound sources were presented from behind a subject. Increasing stimulus duration (from 15-ms to 50-ms) significantly mitigated the rates of front-back confusion and the associated effects of visual stimuli. These findings suggest that concurrent visual stimulation elicits a strong frontal bias in auditory localization and confirm that temporal integration plays an important role in decreasing front-back errors under conditions requiring multisensory spatial processing.

Representation of Multidimensional Stimuli: Quantifying the Most Informative Stimulus Dimension from Neural Responses

A common way to assess the function of sensory neurons is to measure the number of spikes produced by individual neurons while systematically varying a given dimension of the stimulus. Such measured tuning curves can then be used to quantify the accuracy of the neural representation of the stimulus dimension under study, which can in turn be related to behavioral performance. However, tuning curves often change shape when other dimensions of the stimulus are varied, reflecting the simultaneous sensitivity of neurons to multiple stimulus features. Here we illustrate how one-dimensional information analyses are misleading in this context, and propose a framework derived from Fisher information that allows the quantification of information carried by neurons in multidimensional stimulus spaces. We use this method to probe the representation of sound localization in auditory neurons of chinchillas and guinea pigs of both sexes, and show how heterogeneous tuning properties contribute to a representation of sound source position that is robust to changes in sound level.SIGNIFICANCE STATEMENT Sensory neurons' responses are typically modulated simultaneously by numerous stimulus properties, which can result in an overestimation of neural acuity with existing one-dimensional neural information transmission measures. To overcome this limitation, we develop new, compact expressions of Fisher information-derived measures that bound the robust encoding of separate stimulus dimensions in the context of multidimensional stimuli. We apply this method to the problem of the representation of sound source location in the face of changes in sound source level by neurons of the auditory midbrain.

Neural population encoding and decoding of sound source location across sound level in the rabbit inferior colliculus

At lower levels of sensory processing, the representation of a stimulus feature in the response of a neural population can vary in complex ways across different stimulus intensities, potentially changing the amount of feature-relevant information in the response. How higher-level neural circuits could implement feature decoding computations that compensate for these intensity-dependent variations remains unclear. Here we focused on neurons in the inferior colliculus (IC) of unanesthetized rabbits, whose firing rates are sensitive to both the azimuthal position of a sound source and its sound level. We found that the azimuth tuning curves of an IC neuron at different sound levels tend to be linear transformations of each other. These transformations could either increase or decrease the mutual information between source azimuth and spike count with increasing level for individual neurons, yet population azimuthal information remained constant across the absolute sound levels tested (35, 50, and 65 dB SPL), as inferred from the performance of a maximum-likelihood neural population decoder. We harnessed evidence of level-dependent linear transformations to reduce the number of free parameters in the creation of an accurate cross-level population decoder of azimuth. Interestingly, this decoder predicts monotonic azimuth tuning curves, broadly sensitive to contralateral azimuths, in neurons at higher levels in the auditory pathway.

Behavior and modeling of two-dimensional precedence effect in head-unrestrained cats

The precedence effect (PE) is an auditory illusion that occurs when listeners localize nearly coincident and similar sounds from different spatial locations, such as a direct sound and its echo. It has mostly been studied in humans and animals with immobile heads in the horizontal plane; speaker pairs were often symmetrically located in the frontal hemifield. The present study examined the PE in head-unrestrained cats for a variety of paired-sound conditions along the horizontal, vertical, and diagonal axes. Cats were trained with operant conditioning to direct their gaze to the perceived sound location. Stereotypical PE-like behaviors were observed for speaker pairs placed in azimuth or diagonally in the frontal hemifield as the interstimulus delay was varied. For speaker pairs in the median sagittal plane, no clear PE-like behavior occurred. Interestingly, when speakers were placed diagonally in front of the cat, certain PE-like behavior emerged along the vertical dimension. However, PE-like behavior was not observed when both speakers were located in the left hemifield. A Hodgkin-Huxley model was used to simulate responses of neurons in the medial superior olive (MSO) to sound pairs in azimuth. The novel simulation incorporated a low-threshold potassium current and frequency mismatches to generate internal delays. The model exhibited distinct PE-like behavior, such as summing localization and localization dominance. The simulation indicated that certain encoding of the PE could have occurred before information reaches the inferior colliculus, and MSO neurons with binaural inputs having mismatched characteristic frequencies may play an important role.

Segregating two simultaneous sounds in elevation using temporal envelope: Human psychophysics and a physiological model

The ability to segregate simultaneous sound sources based on their spatial locations is an important aspect of auditory scene analysis. While the role of sound azimuth in segregation is well studied, the contribution of sound elevation remains unknown. Although previous studies in humans suggest that elevation cues alone are not sufficient to segregate simultaneous broadband sources, the current study demonstrates they can suffice. Listeners segregating a temporally modulated noise target from a simultaneous unmodulated noise distracter differing in elevation fall into two statistically distinct groups: one that identifies target direction accurately across a wide range of modulation frequencies (MF) and one that cannot identify target direction accurately and, on average, reports the opposite direction of the target for low MF. A non-spiking model of inferior colliculus neurons that process single-source elevation cues suggests that the performance of both listener groups at the population level can be accounted for by the balance of excitatory and inhibitory inputs in the model. These results establish the potential for broadband elevation cues to contribute to the computations underlying sound source segregation and suggest a potential mechanism underlying this contribution.

The precedence effect in sound localization

In ordinary listening environments, acoustic signals reaching the ears directly from real sound sources are followed after a few milliseconds by early reflections arriving from nearby surfaces. Early reflections are spectrotemporally similar to their source signals but commonly carry spatial acoustic cues unrelated to the source location. Humans and many other animals, including nonmammalian and even invertebrate animals, are nonetheless able to effectively localize sound sources in such environments, even in the absence of disambiguating visual cues. Robust source localization despite concurrent or nearly concurrent spurious spatial acoustic information is commonly attributed to an assortment of perceptual phenomena collectively termed "the precedence effect," characterizing the perceptual dominance of spatial information carried by the first-arriving signal. Here, we highlight recent progress and changes in the understanding of the precedence effect and related phenomena.

Linear coding of complex sound spectra by discharge rate in neurons of the medial nucleus of the trapezoid body (MNTB) and its inputs

The interaural level difference (ILD) cue to sound location is first encoded in the lateral superior olive (LSO). ILD sensitivity results because the LSO receives excitatory input from the ipsilateral cochlear nucleus and inhibitory input indirectly from the contralateral cochlear nucleus via glycinergic neurons of the ipsilateral medial nucleus of the trapezoid body (MNTB). It is hypothesized that in order for LSO neurons to encode ILDs, the sound spectra at both ears must be accurately encoded via spike rate by their afferents. This spectral-coding hypothesis has not been directly tested in MNTB, likely because MNTB neurons have been mostly described and studied recently in regards to their abilities to encode temporal aspects of sounds, not spectral. Here, we test the hypothesis that MNTB neurons and their inputs from the cochlear nucleus and auditory nerve code sound spectra via discharge rate. The Random Spectral Shape (RSS) method was used to estimate how the levels of 100-ms duration spectrally stationary stimuli were weighted, both linearly and non-linearly, across a wide band of frequencies. In general, MNTB neurons, and their globular bushy cell inputs, were found to be well-modeled by a linear weighting of spectra demonstrating that the pathways through the MNTB can accurately encode sound spectra including those resulting from the acoustical cues to sound location provided by head-related directional transfer functions (DTFs). Together with the anatomical and biophysical specializations for timing in the MNTB-LSO complex, these mechanisms may allow ILDs to be computed for complex stimuli with rapid spectrotemporally-modulated envelopes such as speech and animal vocalizations and moving sound sources.

Simultaneous comparison of two sound localization measures

Almost all behavioral studies of sound localization have used either an approach-to-target or pointing/orienting task to assess absolute sound localization performance, yet there are very few direct comparisons of these measures. In an approach-to-target task, the subject is trained to walk to a sound source from a fixed location. In an orienting task, finger, head and/or eye movements are monitored while the subject's body is typically constrained. The fact that subjects may also initiate head and eye movements toward the target during the approach-to-target task allows us to measure the accuracy of the initial orienting response and compare it with subsequent target selection. To perform this comparison, we trained cats to localize a broadband noise presented randomly from one of four speakers located ± 30° and ± 60° in azimuth. The cat responded to each sound presentation by walking to and pressing a lever at the perceived location, and a food reward was delivered if the first attempt was correct. In tandem, we recorded initial head and eye orienting movements, via magnetic search coils, immediately following target onset and prior to the walking response. Reducing either stimulus duration or level resulted in a systematic decline in both measurements of localization performance. When the task was easy, localization performance was accurate for both measures. When the task was more difficult, the number of incorrect (i.e., wrong selection) and no-go (i.e., no selection) responses increased. Interestingly, for many of the incorrect trials, there was a dissociation between the orienting response and the target selected, and for many of the no-go trials, the gaze oriented towards the correct target even though the cat did not move to it. This suggests different neural systems governing walking to a target as compared to unconditioned gaze orienting.

Localization of click trains and speech by cats: the negative level effect

Although localization of sound in elevation is believed to depend on spectral cues, it has been shown with human listeners that the temporal features of sound can also greatly affect localization performance. Of particular interest is a phenomenon known as the negative level effect, which describes the deterioration of localization ability in elevation with increasing sound level and is observed only with impulsive or short-duration sound. The present study uses the gaze positions of domestic cats as measures of perceived locations of sound targets varying in azimuth and elevation. The effects of sound level on localization in terms of accuracy, precision, and response latency were tested for sound with different temporal features, such as a click train, a single click, a continuous sound that had the same frequency spectrum of the click train, and speech segments. In agreement with previous human studies, negative level effects were only observed with click-like stimuli and only in elevation. In fact, localization of speech sounds in elevation benefited significantly when the sound level increased. Our findings indicate that the temporal continuity of a sound can affect the frequency analysis performed by the auditory system, and the variation in the frequency spectrum contained in speech sound does not interfere much with the spectral coding for its location in elevation.

Vertical-plane sound localization with distorted spectral cues

For human listeners, the primary cues for localization in the vertical plane are provided by the direction-dependent filtering of the pinnae, head, and upper body. Vertical-plane localization generally is accurate for broadband sounds, but when such sounds are presented at near-threshold levels or at high levels with short durations (<20 ms), the apparent location is biased toward the horizontal plane (i.e., elevation gain <1). We tested the hypothesis that these effects result in part from distorted peripheral representations of sound spectra. Human listeners indicated the apparent position of 100-ms, 50-60 dB SPL, wideband noise-burst targets by orienting their heads. The targets were synthesized in virtual auditory space and presented over headphones. Faithfully synthesized targets were interleaved with targets for which the directional transfer function spectral notches were filled in, peaks were leveled off, or the spectral contrast of the entire profile was reduced or expanded. As notches were filled in progressively or peaks leveled progressively, elevation gain decreased in a graded manner similar to that observed as sensation level is reduced below 30 dB or, for brief sounds, increased above 45 dB. As spectral contrast was reduced, gain dropped only at the most extreme reduction (25% of normal). Spectral contrast expansion had little effect. The results are consistent with the hypothesis that loss of representation of spectral features contributes to reduced elevation gain at low and high sound levels. The results also suggest that perceived location depends on a correlation-like spectral matching process that is sensitive to the relative, rather than absolute, across-frequency shape of the spectral profile.

Effects of forward masking on sound localization in cats: basic findings with broadband maskers

Forward masking is traditionally measured with a detection task in which the addition of a preceding masking sound results in an increased signal-detection threshold. Little is known about the influence of forward masking on localization of free-field sound for human or animal subjects. Here we recorded gaze shifts of two head-unrestrained cats during localization using a search-coil technique. A broadband (BB) noise masker was presented straight ahead. A brief signal could come from 1 of the 17 speaker locations in the frontal hemifield. The signal was either a BB or a band-limited (BL) noise. For BB targets, the presence of the forward masker reduced localization accuracy at almost all target levels (20 to 80 dB SPL) along both horizontal and vertical dimensions. Temporal decay of masking was observed when a 15-ms interstimulus gap was added between the end of the masker and the beginning of the target. A large effect of forward masking was also observed for BL targets with low (0.2-2 kHz) and mid (2-7 kHz) frequencies, indicating that the interaural timing cue is susceptible to forward masking. Except at low sound levels, a small or little effect was observed for high-frequency (7-15 kHz) targets, indicating that the interaural level and the spectral cues in that frequency range remained relatively robust. Our findings suggest that different localization mechanisms can operate independently in a complex listening environment.