Neural rate difference model can account for lateralization of high-frequency stimuli

Integrate-and-fire-type models of the lateral superior olive

Neurons of the lateral superior olive (LSO) in the auditory brainstem play a fundamental role in binaural sound localization. Previous theoretical studies developed various types of neuronal models to study the physiological functions of the LSO. These models were usually tuned to a small set of physiological data with specific aims in mind. Therefore, it is unclear whether and how they can be related to each other, how widely applicable they are, and which model is suitable for what purposes. In this study, we address these questions for six different single-compartment integrate-and-fire (IF) type LSO models. The models are divided into two groups depending on their subthreshold responses: passive (linear) models with only the leak conductance and active (nonlinear) models with an additional low-voltage-activated potassium conductance that is prevalent among the auditory system. Each of these two groups is further subdivided into three subtypes according to the spike generation mechanism: one with simple threshold-crossing detection and voltage reset, one with threshold-crossing detection plus a current to mimic spike shapes, and one with a depolarizing exponential current for spiking. In our simulations, all six models were driven by identical synaptic inputs and calibrated with common criteria for binaural tuning. The resulting spike rates of the passive models were higher for intensive inputs and lower for temporally structured inputs than those of the active models, confirming the active function of the potassium current. Within each passive or active group, the simulated responses resembled each other, regardless of the spike generation types. These results, in combination with the analysis of computational costs, indicate that an active IF model is more suitable than a passive model for accurately reproducing temporal coding of LSO. The simulation of realistic spike shapes with an extended spiking mechanism added relatively small computational costs.

Temporal hyper-precision of brainstem neurons alters spatial sensitivity of binaural auditory processing with cochlear implants

The ability to localize a sound source in complex environments is essential for communication and navigation. Spatial hearing relies predominantly on the comparison of differences in the arrival time of sound between the two ears, the interaural time differences (ITDs). Hearing impairments are highly detrimental to sound localization. While cochlear implants (CIs) have been successful in restoring many crucial hearing capabilities, sound localization via ITD detection with bilateral CIs remains poor. The underlying reasons are not well understood. Neuronally, ITD sensitivity is generated by coincidence detection between excitatory and inhibitory inputs from the two ears performed by specialized brainstem neurons. Due to the lack of electrophysiological brainstem recordings during CI stimulation, it is unclear to what extent the apparent deficits are caused by the binaural comparator neurons or arise already on the input level. Here, we use a bottom-up approach to compare response features between electric and acoustic stimulation in an animal model of CI hearing. Conducting extracellular single neuron recordings in gerbils, we find severe hyper-precision and moderate hyper-entrainment of both the excitatory and inhibitory brainstem inputs to the binaural comparator neurons during electrical pulse-train stimulation. This finding establishes conclusively that the binaural processing stage must cope with highly altered input statistics during CI stimulation. To estimate the consequences of these effects on ITD sensitivity, we used a computational model of the auditory brainstem. After tuning the model parameters to match its response properties to our physiological data during either stimulation type, the model predicted that ITD sensitivity to electrical pulses is maintained even for the hyper-precise inputs. However, the model exhibits severely altered spatial sensitivity during electrical stimulation compared to acoustic: while resolution of ITDs near midline was increased, more lateralized adjacent source locations became inseparable. These results directly resemble recent findings in rodent and human CI listeners. Notably, decreasing the phase-locking precision of inputs during electrical stimulation recovered a wider range of separable ITDs. Together, our findings suggest that a central problem underlying the diminished ITD sensitivity in CI users might be the temporal hyper-precision of inputs to the binaural comparator stage induced by electrical stimulation.

Contextual Lateralization Based on Interaural Level Differences Is Preshaped by the Auditory Periphery and Predominantly Immune Against Sequential Segregation

The perceived azimuth of a target sound is determined by the interaural time difference and the interaural level difference (ILD) and is subject to contextual effects from precursor sounds. This study characterized ILD-based precursor effects (PEs) for high-frequency stimuli in a total of seven normal-hearing listeners. In Experiment 1, precursor and target were band-pass-filtered noises approximately centered at 4 kHz (1.2- and 1-octave bandwidth, respectively) separated by a 10-ms gap. The effects of precursor location (ipsilateral, contralateral, and central) on the perceived target azimuth were measured using a head-pointing task. Relative to control trials without a precursor, ipsilateral precursors biased the perceived target azimuth toward midline (medial bias) and contralateral precursors biased it contralaterally (lateral bias). Central precursors caused a symmetric lateral bias. An auditory periphery model that determines the "internal" ILD at the auditory nerve level, including either realistic efferent compression control or auditory nerve adaptation, explained about 50% of the variance in the PEs. These within-trial PEs were accompanied by an across-trial PE, inducing medial bias. Experiment 2 studied the role of sequential segregation in the within-trial PE by introducing a pitch difference between precursor and target. Segregation conditions caused increased PE for ipsilateral, no effect for contralateral, and either no effect or reduced PE for central precursors. Overall, the ILD-based within-trial PE appears to be preshaped already in the auditory periphery and the mechanism underlying at least the ipsilateral PE appears to be immune against sequential segregation.

A Computational Model of a Single Auditory Nerve Fiber for Electric-Acoustic Stimulation

Cochlear implant (CI) recipients with preserved acoustic low-frequency hearing in the implanted ear are a growing group among traditional CI users who benefit from hybrid electric-acoustic stimulation (EAS). However, combined ipsilateral electric and acoustic stimulation also introduces interactions between the two modalities that can affect the performance of EAS users. A computational model of a single auditory nerve fiber that is excited by EAS was developed to study the interaction between electric and acoustic stimulation. Two existing models of sole electric or acoustic stimulation were coupled to simulate responses to combined EAS. Different methods of combining both models were implemented. In the coupled model variant, the refractoriness of the simulated fiber leads to suppressive interaction between electrically evoked and acoustically evoked spikes as well as spontaneous activity. The second model variant is an uncoupled EAS model without electric-acoustic interaction. By comparing predictions between the coupled and the noninteracting EAS model, it was possible to infer electric-acoustic interaction at the level of the auditory nerve. The EAS model was used to simulate fiber populations with realistic inter-unit variability, where each unit was represented by the single-fiber model. Predicted thresholds and dynamic ranges, spike rates, latencies, jitter, and vector strengths were compared to empirical data. The presented EAS model provides a framework for future studies of peripheral electric-acoustic interaction.

A comparative study of eight human auditory models of monaural processing

A number of auditory models have been developed using diverging approaches, either physiological or perceptual, but they share comparable stages of signal processing, as they are inspired by the same constitutive parts of the auditory system. We compare eight monaural models that are openly accessible in the Auditory Modelling Toolbox. We discuss the considerations required to make the model outputs comparable to each other, as well as the results for the following model processing stages or their equivalents: Outer and middle ear, cochlear filter bank, inner hair cell, auditory nerve synapse, cochlear nucleus, and inferior colliculus. The discussion includes a list of recommendations for future applications of auditory models.

Simulation of ITD-Dependent Single-Neuron Responses Under Electrical Stimulation and with Amplitude-Modulated Acoustic Stimuli

Interaural time difference (ITD) sensitivity with cochlear implant stimulation is remarkably similar to envelope ITD sensitivity using conventional acoustic stimulation. This holds true for human perception, as well as for neural response rates recorded in the inferior colliculus of several mammalian species. We hypothesize that robust excitatory-inhibitory (EI) interaction is the dominant mechanism. Therefore, we connected the same single EI-model neuron to either a model of the normal acoustic auditory periphery or to a model of the electrically stimulated auditory nerve. The model captured most features of the experimentally obtained response properties with electric stimulation, such as the shape of rate-ITD functions, the dependence on stimulation level, and the pulse rate or modulation-frequency dependence. Rate-ITD functions with high-rate, amplitude-modulated electric stimuli were very similar to their acoustic counterparts. Responses obtained with unmodulated electric pulse trains most resembled acoustic filtered clicks. The fairly rapid decline of ITD sensitivity at rates above 300 pulses or cycles per second is correctly simulated by the 3.1-ms time constant of the inhibitory post-synaptic conductance. As the model accounts for these basic properties, it is expected to help in understanding and quantifying the binaural hearing abilities with electric stimulation when integrated in bigger simulation frameworks.

Sensitivity to Envelope Interaural Time Differences: Modeling Auditory Modulation Filtering

For amplitude-modulated sound, the envelope interaural time difference (ITD_ENV) is a potential cue for sound-source location. ITD_ENV is encoded in the lateral superior olive (LSO) of the auditory brainstem, by excitatory-inhibitory (EI) neurons receiving ipsilateral excitation and contralateral inhibition. Between human listeners, sensitivity to ITD_ENV varies considerably, but ultimately decreases with increasing stimulus carrier frequency, and decreases more strongly with increasing modulation rate. Mechanisms underlying the variation in behavioral sensitivity remain unclear. Here, with increasing carrier frequency (4-10 kHz), as we phenomenologically model the associated decrease in ITD_ENV sensitivity using arbitrarily fewer neurons consistent across populations, we computationally model the variable sensitivity across human listeners and modulation rates (32-800 Hz) as the decreasing range of membrane frequency responses in LSO neurons. Transposed tones stimulate a bilateral auditory-periphery model, driving model EI neurons where electrical membrane impedance filters the frequency content of inputs driven by amplitude-modulated sound, evoking modulation filtering. Calculated from Fisher information in spike-rate functions of ITD_ENV, for model EI neuronal populations distinctly reflecting the LSO range in membrane frequency responses, just-noticeable differences in ITD_ENV collectively reproduce the largest variation in ITD_ENV sensitivity across human listeners. These slow to fast model populations each generally match the best human ITD_ENV sensitivity at a progressively higher modulation rate, by membrane-filtering and spike-generation properties producing realistically less than Poisson variance. Non-resonant model EI neurons are also sensitive to interaural intensity differences. With peripheral filters centered between carrier frequency and modulation sideband, fast resonant model EI neurons extend ITD_ENV sensitivity above 500-Hz modulation.

Influence of envelope fluctuation on the lateralization of interaurally delayed low-frequency stimuli

Disregarding onset and offset effects, interaurally delaying a 500 Hz tone by 1.5 ms is identical to advancing it by 0.5 ms. When presented over headphones, humans indeed perceive such a tone lateralized toward the side of the nominal lag. Any stimulus other than a tone has more than one frequency component and is thus unambiguous. It has been shown that phase ambiguity can be resolved when increasing the stimulus bandwidth. This has mostly been attributed to the integration of information across frequencies. Additionally, interaural timing information conveyed in the stimulus envelope within a single frequency channel is a second possible cue that could help to resolve phase ambiguity. This study employs stimuli designed to differ in the amount of envelope fluctuation while retaining the same power spectral density as well as interaural differences. Any difference in lateralization must thus be a result of the difference in envelope. The results show that stimuli with strong envelope fluctuation require significantly smaller bandwidths to resolve phase ambiguity when compared to stimuli with weak envelope fluctuation. This suggests that within-channel information is an important cue used to resolve phase ambiguity.

Robustness of neuronal tuning to binaural sound localization cues against age-related loss of inhibitory synaptic inputs

Sound localization relies on minute differences in the timing and intensity of sound arriving at both ears. Neurons of the lateral superior olive (LSO) in the brainstem process these interaural disparities by precisely detecting excitatory and inhibitory synaptic inputs. Aging generally induces selective loss of inhibitory synaptic transmission along the entire auditory pathways, including the reduction of inhibitory afferents to LSO. Electrophysiological recordings in animals, however, reported only minor functional changes in aged LSO. The perplexing discrepancy between anatomical and physiological observations suggests a role for activity-dependent plasticity that would help neurons retain their binaural tuning function despite loss of inhibitory inputs. To explore this hypothesis, we use a computational model of LSO to investigate mechanisms underlying the observed functional robustness against age-related loss of inhibitory inputs. The LSO model is an integrate-and-fire type enhanced with a small amount of low-voltage activated potassium conductance and driven with (in)homogeneous Poissonian inputs. Without synaptic input loss, model spike rates varied smoothly with interaural time and level differences, replicating empirical tuning properties of LSO. By reducing the number of inhibitory afferents to mimic age-related loss of inhibition, overall spike rates increased, which negatively impacted binaural tuning performance, measured as modulation depth and neuronal discriminability. To simulate a recovery process compensating for the loss of inhibitory fibers, the strength of remaining inhibitory inputs was increased. By this modification, effects of inhibition loss on binaural tuning were considerably weakened, leading to an improvement of functional performance. These neuron-level observations were further confirmed by population modeling, in which binaural tuning properties of multiple LSO neurons were varied according to empirical measurements. These results demonstrate the plausibility that homeostatic plasticity could effectively counteract known age-dependent loss of inhibitory fibers in LSO and suggest that behavioral degradation of sound localization might originate from changes occurring more centrally.