Objective Estimation of Sensory Thresholds Based on Neurophysiological Parameters

Predictive coding and stochastic resonance as fundamental principles of auditory phantom perception

Mechanistic insight is achieved only when experiments are employed to test formal or computational models. Furthermore, in analogy to lesion studies, phantom perception may serve as a vehicle to understand the fundamental processing principles underlying healthy auditory perception. With a special focus on tinnitus-as the prime example of auditory phantom perception-we review recent work at the intersection of artificial intelligence, psychology and neuroscience. In particular, we discuss why everyone with tinnitus suffers from (at least hidden) hearing loss, but not everyone with hearing loss suffers from tinnitus. We argue that intrinsic neural noise is generated and amplified along the auditory pathway as a compensatory mechanism to restore normal hearing based on adaptive stochastic resonance. The neural noise increase can then be misinterpreted as auditory input and perceived as tinnitus. This mechanism can be formalized in the Bayesian brain framework, where the percept (posterior) assimilates a prior prediction (brain's expectations) and likelihood (bottom-up neural signal). A higher mean and lower variance (i.e. enhanced precision) of the likelihood shifts the posterior, evincing a misinterpretation of sensory evidence, which may be further confounded by plastic changes in the brain that underwrite prior predictions. Hence, two fundamental processing principles provide the most explanatory power for the emergence of auditory phantom perceptions: predictive coding as a top-down and adaptive stochastic resonance as a complementary bottom-up mechanism. We conclude that both principles also play a crucial role in healthy auditory perception. Finally, in the context of neuroscience-inspired artificial intelligence, both processing principles may serve to improve contemporary machine learning techniques.

Objective hearing threshold identification from auditory brainstem response measurements using supervised and self-supervised approaches

Hearing loss is a major health problem and psychological burden in humans. Mouse models offer a possibility to elucidate genes involved in the underlying developmental and pathophysiological mechanisms of hearing impairment. To this end, large-scale mouse phenotyping programs include auditory phenotyping of single-gene knockout mouse lines. Using the auditory brainstem response (ABR) procedure, the German Mouse Clinic and similar facilities worldwide have produced large, uniform data sets of averaged ABR raw data of mutant and wildtype mice. In the course of standard ABR analysis, hearing thresholds are assessed visually by trained staff from series of signal curves of increasing sound pressure level. This is time-consuming and prone to be biased by the reader as well as the graphical display quality and scale.In an attempt to reduce workload and improve quality and reproducibility, we developed and compared two methods for automated hearing threshold identification from averaged ABR raw data: a supervised approach involving two combined neural networks trained on human-generated labels and a self-supervised approach, which exploits the signal power spectrum and combines random forest sound level estimation with a piece-wise curve fitting algorithm for threshold finding.We show that both models work well and are suitable for fast, reliable, and unbiased hearing threshold detection and quality control. In a high-throughput mouse phenotyping environment, both methods perform well as part of an automated end-to-end screening pipeline to detect candidate genes for hearing involvement. Code for both models as well as data used for this work are freely available.

Classification at the accuracy limit: facing the problem of data ambiguity

Data classification, the process of analyzing data and organizing it into categories or clusters, is a fundamental computing task of natural and artificial information processing systems. Both supervised classification and unsupervised clustering work best when the input vectors are distributed over the data space in a highly non-uniform way. These tasks become however challenging in weakly structured data sets, where a significant fraction of data points is located in between the regions of high point density. We derive the theoretical limit for classification accuracy that arises from this overlap of data categories. By using a surrogate data generation model with adjustable statistical properties, we show that sufficiently powerful classifiers based on completely different principles, such as perceptrons and Bayesian models, all perform at this universal accuracy limit under ideal training conditions. Remarkably, the accuracy limit is not affected by certain non-linear transformations of the data, even if these transformations are non-reversible and drastically reduce the information content of the input data. We further compare the data embeddings that emerge by supervised and unsupervised training, using the MNIST data set and human EEG recordings during sleep. We find for MNIST that categories are significantly separated not only after supervised training with back-propagation, but also after unsupervised dimensionality reduction. A qualitatively similar cluster enhancement by unsupervised compression is observed for the EEG sleep data, but with a very small overall degree of cluster separation. We conclude that the handwritten letters in MNIST can be considered as 'natural kinds', whereas EEG sleep recordings are a relatively weakly structured data set, so that unsupervised clustering will not necessarily re-cover the human-defined sleep stages.

Preventive Effects of Ginkgo-Extract EGb 761® on Noise Trauma-Induced Cochlear Synaptopathy

Noise trauma-induced loss of ribbon synapses at the inner hair cells (IHC) of the cochlea may lead to hearing loss (HL), resulting in tinnitus. We are convinced that a successful and sustainable therapy of tinnitus has to treat both symptom and cause. One of these causes may be the mentioned loss of ribbon synapses at the IHC of the cochlea. In this study, we investigated the possible preventive and curative effects of the Ginkgo biloba extract EGb 761® on noise-induced synaptopathy, HL, and tinnitus development in Mongolian gerbils (Meriones unguiculatus). To this end, 37 male animals received EGb 761® or placebo orally 3 weeks before (16 animals) or after (21 animals) a monaural acoustic noise trauma (2 kHz, 115 dB SPL, 75 min). Animals' hearing thresholds were determined by auditory brainstem response (ABR) audiometry. A possible tinnitus percept was assessed by the gap prepulse inhibition acoustic startle reflex (GPIAS) response paradigm. Synaptopathy was quantified by cochlear immunofluorescence histology, counting the ribbon synapses of 15 IHCs at 11 different cochlear frequency locations per ear. We found a clear preventive effect of EGb 761® on ribbon synapse numbers with the surprising result of a significant increase in synaptic innervation on the trauma side relative to placebo-treated animals. Consequently, animals treated with EGb 761® before noise trauma did not develop a significant HL and were also less affected by tinnitus compared to placebo-treated animals. On the other hand, we did not see a curative effect (EGb 761® treatment after noise trauma) of the extract on ribbon synapse numbers and, consequently, a significant HL and no difference in tinnitus development compared to the placebo-treated animals. Taken together, EGb 761® prevented noise-induced HL and tinnitus by protecting from noise trauma-induced cochlear ribbon synapse loss; however, in our model, it did not restore lost ribbon synapses.

Circadian Sensitivity of Noise Trauma-Induced Hearing Loss and Tinnitus in Mongolian Gerbils

Noise-induced hearing loss (HL) has a circadian component: In nocturnal mice, hearing thresholds (HT) have a significantly stronger effect to acoustic trauma when induced during the night compared to rather mild effects on hearing when induced during daytime. Here, we investigate whether such effects are also present in diurnal Mongolian gerbils and determined whether trauma-induced HL correlated with the development of a tinnitus percept in these animals. In particular, we investigated the effects of acoustic trauma (2 kHz, 115 dB SPL, 75 min) on HT and tinnitus development in 34 male gerbils exposed either at 9 AM, 1 PM, 5 PM, or 12 PM. HT was measured by acoustic brainstem response audiometry at defined times 1 day before and 1 week after the trauma. Possible tinnitus percepts were assessed behaviorally by the gap prepulse inhibition of the acoustic startle response at defined times 1 day before and 1 week after the trauma. We found daytime-dependent changes due to trauma in mean HT in a frequency-dependent manner comparable to the results in mice, but the results temporally shifted according to respective activity profiles. Additionally, we found linear correlations of these threshold changes with the strength of the tinnitus percept, with the most prominent correlations in the 5 PM trauma group. Taken together, circadian sensitivity of the HT to noise trauma can also be found in gerbils, and tinnitus strength correlates most strongly with HL only when the trauma is applied at the most sensitive times, which seem to be the evening.

Real-time threshold determination of auditory brainstem responses by cross-correlation analysis

Auditory brainstem response (ABR) serves as an objective indication of auditory perception at a given sound level and is nowadays widely used in hearing function assessment. Despite efforts for automation over decades, ABR threshold determination by machine algorithms remains unreliable and thereby one still relies on visual identification by trained personnel. Here, we described a procedure for automatic threshold determination that can be used in both animal and human ABR tests. The method terminates level averaging of ABR recordings upon detection of time-locked waveform through cross-correlation analysis. The threshold level was then indicated by a dramatic increase in the sweep numbers required to produce “qualified” level averaging. A good match was obtained between the algorithm outcome and the human readouts. Moreover, the method varies the level averaging based on the cross-correlation, thereby adapting to the signal-to-noise ratio of sweep recordings. These features empower a robust and fully automated ABR test.

•

Automatic threshold determination of auditory brainstem response (ABR)

•

Detection of “clear” responses from iteratively averaged level representation

•

Wide application in both animal and human ABR tests

•

Stop on-going level averaging based on detection outcome

Salicylate-Induced Changes in Hearing Thresholds in Mongolian Gerbils Are Correlated With Tinnitus Frequency but Not With Tinnitus Strength

Tinnitus is an auditory phantom percept without external sound sources. Despite the high prevalence and tinnitus-associated distress of affected patients, the pathophysiology of tinnitus remains largely unknown, making prevention and treatments difficult to develop. In order to elucidate the pathophysiology of tinnitus, animal models are used where tinnitus is induced either permanently by noise trauma or transiently by the application of salicylate. In a model of trauma-induced tinnitus, we have suggested a central origin of tinnitus-related development of neuronal hyperactivity based on stochastic resonance (SR). SR refers to the physiological phenomenon that weak subthreshold signals for given sensors (or synapses) can still be detected and transmitted if appropriate noise is added to the input of the sensor. The main objective of this study was to characterize the neurophysiological and behavioral effects during salicylate-induced tinnitus and compare these to the conditions within the trauma model. Our data show, in line with the pharmacokinetics, that hearing thresholds generally increase 2 h after salicylate injections. This increase was significantly stronger within the region of best hearing compared to other frequencies. Furthermore, animals showed behavioral signs of tinnitus during that time window and frequency range as assessed by gap prepulse inhibition of the acoustic startle reflex (GPIAS). In contrast to animals with noise trauma-induced tinnitus, salicylate-induced tinnitus animals showed no correlation between hearing thresholds and behavioral signs of tinnitus, indicating that the development of tinnitus after salicylate injection is not based on SR as proposed for the trauma model. In other words, salicylate-induced tinnitus and noise trauma-induced tinnitus are not based on the same neurophysiological mechanism.

Simulated transient hearing loss improves auditory sensitivity

Recently, it was proposed that a processing principle called adaptive stochastic resonance plays a major role in the auditory system, and serves to maintain optimal sensitivity even to highly variable sound pressure levels. As a side effect, in case of reduced auditory input, such as permanent hearing loss or frequency specific deprivation, this mechanism may eventually lead to the perception of phantom sounds like tinnitus or the Zwicker tone illusion. Using computational modeling, the biological plausibility of this processing principle was already demonstrated. Here, we provide experimental results that further support the stochastic resonance model of auditory perception. In particular, Mongolian gerbils were exposed to moderate intensity, non-damaging long-term notched noise, which mimics hearing loss for frequencies within the notch. Remarkably, the animals developed significantly increased sensitivity, i.e. improved hearing thresholds, for the frequency centered within the notch, but not for frequencies outside the notch. In addition, most animals treated with the new paradigm showed identical behavioral signs of phantom sound perception (tinnitus) as animals with acoustic trauma induced tinnitus. In contrast, animals treated with broadband noise as a control condition did not show any significant threshold change, nor behavioral signs of phantom sound perception.

Tinnitus development is associated with synaptopathy of inner hair cells in Mongolian gerbils

Human hearing loss (HL) is often accompanied by comorbidities like tinnitus, which is affecting up to 15% of the adult population. Rodent animal studies could show that tinnitus may not only be a result of apparent HL due to cochlear hair cell damage but can also be a consequence of synaptopathy at the inner hair cells (IHCs) already induced by moderate sound traumata. Here, we investigate synaptopathy previously shown in mice in our animal model, the Mongolian gerbil, and relate it to behavioral signs of tinnitus. Tinnitus was induced by a mild monaural acoustic trauma leading to monaural noise induced HL in the animals, quantified by auditory brainstem response (ABR) audiometry. Behavioral signs of tinnitus percepts were detected by measurement of prepulse inhibition of the acoustic startle response in a gap-noise paradigm. Fourteen days after trauma, the cochleae of both ears were isolated, and IHC synapses were counted within several spectral regions of the cochlea. Behavioral signs of tinnitus were only found in animals with IHC synaptopathy, independent of type of HL. On the other hand, animals with apparent HL but without behavioral signs of tinnitus showed a reduction in amplitudes of ABR waves I&II but no significant changes in the number of synapses at the IHC. We conclude-in line with the literature-that HL is caused by damage to the IHC or by other reasons but that the development of tinnitus, at least in our animal model, is closely linked to synaptopathy at the IHC.

Open(G)PIAS: An Open-Source Solution for the Construction of a High-Precision Acoustic Startle Response Setup for Tinnitus Screening and Threshold Estimation in Rodents

The modulation of the acoustic startle reflex (ASR) by a pre-stimulus called pre-pulse inhibition (PPI, for gap of silence pre-stimulus: GPIAS) is a versatile tool to, e.g., estimate hearing thresholds or identify subjective tinnitus percepts in rodents. A proper application of these paradigms depends on a reliable measurement of the ASR amplitudes and an exact stimulus presentation in terms of frequency and intensity. Here, we introduce a novel open-source solution for the construction of a low-cost ASR setup. The complete software for data acquisition and stimulus presentation is written in Python 3.6 and is provided as an Anaconda package. Furthermore, we provide a construction plan for the sensor system based on low-cost hardware components. Exemplary GPIAS data from two animal models (Mus musculus, Meriones unguiculatus) show that the ratio histograms (1-GPIAS) of the gap-pre-stimulus and no pre-stimulus ASR amplitudes can be well described by a log-normal distribution being in good accordance to previous studies with already established setups. Furthermore, it can be shown that the PPI as a function of pre-stimulus intensity (threshold paradigm) can be approximated with a hard-sigmoid function enabling a reproducible sensory threshold estimation. Thus, we show that the open-source solution could help to further establish the ASR method in many laboratories and, thus, facilitate and standardize research in animal models of tinnitus and/or hearing loss.

Objective Estimation of Sensory Thresholds Based on Neurophysiological Parameters

Reliable determination of sensory thresholds is the holy grail of signal detection theory. However, there exists no assumption-independent gold standard for the estimation of thresholds based on neurophysiological parameters, although a reliable estimation method is crucial for both scientific investigations and clinical diagnosis. Whenever it is impossible to communicate with the subjects, as in studies with animals or neonates, thresholds have to be derived from neural recordings or by indirect behavioral tests. Whenever the threshold is estimated based on such measures, the standard approach until now is the subjective setting-either by eye or by statistical means-of the threshold to the value where at least a "clear" signal is detectable. These measures are highly subjective, strongly depend on the noise, and fluctuate due to the low signal-to-noise ratio near the threshold. Here we show a novel method to reliably estimate physiological thresholds based on neurophysiological parameters. Using surrogate data we demonstrate that fitting the responses to different stimulus intensities with a hard sigmoid function, in combination with subsampling, provides a robust threshold value as well as an accurate uncertainty estimate. This method has no systematic dependence on the noise and does not even require samples in the full dynamic range of the sensory system. We prove that this method is universally applicable to all types of sensory systems, ranging from somatosensory stimulus processing in the cortex to auditory processing in the brain stem.