2017 Seminars
January 2017
| HRC Seminar with Elizabeth Casserly | January 20th |
Elizabeth Casserly, Trinity College
Title: Auditory training using TV interviews: Harnessing attention and top-down cognitive networks
Abstract: A major effort has been undertaken in recent years to develop better training methods for improving auditory perception in hearing-impaired populations and cochlear implant users. These individuals vary widely in their functional hearing capacity, and training protocols based on improvements in low-level discrimination ability and sentence-level word recognition have had limited success in producing real-world perceptual gains. In this talk I will describe the ongoing effort in my lab to improve training generalization and real-world benefits by using naturalistic training materials. We have transcribed and edited 15 interview segments featuring three well-known hosts (Ellen DeGeneres, David Letterman, and Zach Galifianakis). During training participants alternate between passive listening and utterance transcription with orthographic feedback. The method provides a number of potential benefits: increased attention and engagement, knowledge of familiar voices, availability of top-down discourse-level semantic cues for speech perception, and exposure to multiple talkers. Initial tests with simulated CI signals and normal-hearing college students have been promising, but there is still room for improvement. I will conclude by suggesting further ways in which positive cognitive influences on learning may be applied to the auditory training method. We hope to take advantage of some of these areas, expand the training set, and examine the effects in clinical populations in future work.
| HRC Seminar with Dan Sanes | January 27th |
Dan Sanes, New York University
Title: Influence of early experience on sensory and non-sensory processing
Abstract: Even a transient period of hearing loss can induce deficits in auditory perception and aural communication skills, when it occurs during childhood. One explanation for this is that developmental hearing loss causes irreversible changes to the developing nervous system, thereby degrading central auditory processing. I will present evidence showing that hearing loss-induced impairments to auditory perception and neural encoding are associated with synaptic and biophysical changes within auditory cortex that persist into adulthood. We have also found that developmental hearing loss is a risk factor for delayed learning on auditory tasks. One plausible basis for learning deficits is that hearing loss causes long-lasting impairments to brain areas downstream of auditory cortex that are commonly associated with cognitive abilities. In fact, following hearing loss, synaptic transmission fails to develop properly in auditory striatum, a downstream target of auditory cortex that is associated with reward learning. Taken together, our results suggest that a period of developmental hearing loss can derail both sensory and non-sensory neural mechanisms, providing an explanation for the barriers to auditory perception or learning that can persist long after normal audibility is restored.
February 2017
| HRC Seminar with Heather Reed | February 17th |
Heather Reed, University of Connecticut
Title: Brain Organization for Discriminating Ears
Abstract: Mammals discriminate sounds based on perception of temporal cues in the sound envelope and they need cortex to detect these cues but the underlying cortical circuits and coding mechanisms remain unknown. First, I will summarize how we have quantified the temporal cues in natural vocalization sequences of various species in order to identify universal versus species-specific features. To explore neural coding of these cues, we first locate primary and non-primary cortices in the rat with wide-field imaging. Single neuron response spike trains are then recorded from layer 4 neurons of primary (A1) and ventral non-primary cortices (VAF and SRAF) of the rat as the physiology and corresponding thalamocortical pathways have been well described (Polley et al., 2007, Storace et al., 2010, 2011, 2012, Higgins et al., 2010). Next, spiking responses are probed with a series of synthetic noise burst sequences that have temporal cues matching those of natural vocalizations. We find A1 has primarily brief responses to sound onset whereas non-primary cortices have more sustained responses. Moreover, within each cortical field the response duration increases in proportion to the sound duration creating a unique neural code for each unique noise burst sequence. Next we explored whether these spike timing sequences could be used to discriminate sounds and find that indeed they can. Furthermore, non-primary cortices require longer time scales for optimal sound sequence discrimination giving them the capacity to group acoustic features over longer time scales than primary cortex. The multi-scaled organization could in theory provide a behavioral advantage to mammals for discriminating temporal shape cues in the sound envelope.
March 2017
| HRC Seminar with Mark Riggle | March 24th |
Mark Riggle
Title: The Vestibular Origins of Airborne Hearing Determined the Evolutionary Path for Enabling Spatial Hearing and for Developing Auditory Specializations
Abstract: Evolutionary constraints on the origins of hearing suggest an alternative neural architecture for spatial hearing. One constraint is that land animals were severely deaf prior to the development of tympanic membrane hearing (TMH) which occurred about 100 million years after mammals and archosaurs (bird ancestors) had separated. Before TMH, because any sound perception could only arise from skull vibrations, there could be no sensory specialization for hearing (a cochlea), no ITD, no ILD, and no spectral information. Additionally, because the cochlea descended from a vestibular otolith organ and because TMH would have caused sounds to produce vibrations in the vestibular fluids, that therefore, at the origin of TMH, external sounds would have been encoded in the vestibular neural signals. However, because evolution requires that TMH be immediately useful, initial hearing must have occurred then and without any auditory neural specialization. We show how spatial hearing directly emerged at the origin of TMH by reuse of other neural systems (specifically visual orientation, cerebellar vestibular processing, and tactile somatosensory localization). From this initial hearing capability, evolution developed the cochlea and the neural specializations of the ascending auditory pathway — surprisingly, that pathway may not function for localization but rather became specialized for multiple-source separation.
We show how initial sound localization arises from the following:
1. Sound sources can be accurately localized by a simple mathematical operator;
2. By applying the operator for sound localization, the sound signal encoded on the vestibular signal can be spatially localized by the vestibular/oculomotor cerebellum;
3. Tactile localization (ancient back to fish) reuses parts of the visual saccade system — a system which is now known to need the oculomotor cerebellum (OMC) for both learning how and for performing accurate saccades.
4. Spatial hearing then began with TMH by reusing the tactile localization system with the OMC, and that provided the hearing substrate for evolution to work with.
| HRC Seminar with Kirill Nourski | March 31st |
Kirill Nourski, University of Iowa
Title: Functional organization of human auditory cortex: An intracranial electrophysiology perspective
Abstract: Human auditory cortex is structurally complex and challenging to study. Current models of primate auditory cortex organization delineate it into core, belt, and parabelt fields, yet the identity, location and functional specialization of these fields in humans remains unclear. Intracranial electrophysiological recordings in epilepsy patients offer an opportunity to study human auditory cortical processing with unprecedented spatiotemporal resolution. Better understanding of basic electrophysiological properties of human auditory cortex is achieved through the systematic study of responses to simple and complex sounds including speech, characterization of onset response latencies and effects of general anesthesia on auditory cortical activity. This presentation summarizes recent intracranial studies of human auditory cortex with an emphasis on basic response properties and modulation of cortical activity during induction of general anesthesia and performance of active behavioral tasks. The tonotopically organized core auditory cortex represents acoustic attributes of sounds with high temporal precision and short response latencies; activity there is minimally modulated by task, attention or general anesthesia. While non-core cortex also represents stimulus acoustic features, its activity is modulated by task requirements and is abolished upon induction of anesthesia. Finally, responses within auditory-related prefrontal areas are strongly modulated by task requirements and correlate with behavioral performance on speech identification tasks. Direct recordings thus reveal a hierarchical organization of sound processing within human auditory and auditory-related cortex.
April 2017
| HRC Seminar with David Poeppel | April 7th |
David Poeppel, New York University
Title: Speech is special and language is structured
Abstract: I discuss recent (quite fun and straightforward) experiments that focus on general questions about the cognitive science and neural implementation of speech and language. I come to (currently) unpopular conclusions about both domains. Based on a first set of experiments, using fMRI and exploiting the temporal statistics of speech, I argue for the existence of a speech-specific processing stage that implicates a particular neuronal substrate (the superior temporal sulcus) that has the appropriate sensitivity and selectivity for speech. Based on a second set of experiments, using MEG, I outline neural mechanisms that can form the basis for more abstract, structural processing. The results demonstrate that, during listening to connected speech, cortical activity at different time scales is entrained concurrently to track the time course of linguistic structures at different hierarchical levels. The results demonstrate syntax-driven, internal construction of hierarchical linguistic structure via entrainment of hierarchical cortical dynamics. The conclusions — that speech is special and language syntactic-structure-driven — provide neurobiological provocations to the prevailing view that speech perception is ‘mere’ hearing and that language comprehension is ‘mere’ statistics.
| HRC Seminar with Josh McDermott | April 14th |
Josh McDermott, MIT
Title: Computational Neuroimaging of Human Auditory Cortex
Abstract: Just by listening, humans can determine who is talking to them, whether a window in their house is open or shut, or what their child dropped on the floor in the next room. This ability to derive information from sound is enabled by a cascade of neuronal processing stages that transform the sound waveform entering the ear into cortical representations that are presumed to make behaviorally important sound properties explicit. Although much is known about the peripheral processing of sound, the auditory cortex remains poorly understood, with little consensus even about its coarse-scale organization in humans. This talk will describe our recent efforts using computational neuroimaging methods to better understand the cortical representation of sound. I will describe the use of “model-matched” stimuli to test whether existing models explain cortical neural responses, and the development of new models of cortical computation via task-optimization of deep neural networks. We have harnessed these methods and models to reveal representational transformations occurring between primary and non-primary cortex that may support the recognition of speech, music, and other real-world sound signals.
| HRC Seminar with Heath Jones | April 21st |
Heath Jones, United States Army Aeromedical Research Laboratory
Title: Lasers and Choppers, and Ears, oh my!
Abstract: The talk will include discussions of hearing protection research efforts at the U.S. Army Aeromedical Research Laboratory (USAARL), including assessing the risk of auditory injury to impulsive noise and to aircraft noise on active airfields. Recently, the United States Department of Defense acquisition standard (MIL-STD-1474E) was adopted that mandates the U.S. Army use the Auditory Hazard Assessment Algorithm for Humans (AHAAH) for calculating maximum permissible noise levels and exposure limits of military systems. The AHAAH is an electro-acoustic model designed to predict the auditory injury that results from intense pressure changes at the ear caused by blast overpressure and acoustic impulsive noise exposures. However, there are concerns about the appropriateness of implementing it as a medical standard in an updated Damage Risk Criterion. One concern involves the inclusion of a “warned” option in the AHAAH for calculating the risk factor metric for a given impulse. The model developers suggest that extended experience can classically condition the acoustic reflex to be activated at the time of impulse arrival, which provides protection and reduces the risk factor for auditory injury. To date, there is limited evidence of acoustic reflex classical conditioning. For the purpose of a health hazard assessment, an inappropriate implementation of this assumption would result in an underestimate of auditory hazard and may incorrectly predict that some high-level exposures are safe. Ongoing research at USAARL that was developed to directly test these issues will be discussed. In addition, recent noise measurements around the UH-60 Blackhawk will be presented. Hearing is critical to the performance of personnel and integral to speech communications during airfield operations. Aircraft engines and subsystems are major sources of noise, which impairs communication between pilots, ground crews, and on/off boarding passengers. One population of military personnel routinely exposed to hazardous noise conditions includes aircraft ground support and maintenance crews. As it is not possible to reduce the source of the aircraft noise for these personnel, protection from excessive noise levels, as well as promotion of communication abilities and situational awareness are important for increasing safety to military personnel on airfields. This presentation will detail current methods for assessing hearing health risk associated with impulsive noises and efforts for reducing risk associated with steady-state noises on active airfields.
| HRC Seminar with Darlene Ketten | April 28th |
Darlene Ketten, Boston University
Title: Monty Python Hearing: Functional analyses of something different in land and marine micro and mega ears
Abstract: Ultrasonic echolocation (biosonar) is a well-documented ability of most bats (Microchiroptera) and toothed whales (Cetacea: Odontoceti) whereas elephants (Proboscidea) and baleen whales (Cetacea: Mysticeti) are massive creatures that respond to exceptionally low frequency signals. This raises two basic questions. First, these animals operate in different media, air vs. water, with radically different acoustic properties: 4.5-fold differences in sound speed; three magnitudes in acoustic impedance. Further, a commonly held tenet is that the upper limit of hearing is inversely correlated with body mass, implying there should be little or no overlap in hearing across these taxa. However, data from audiometric and vocalization studies on these groups suggest there are similarities in their hearing abilities despite differences in habitat and size.
There are also practical issues driving research on hearing in these animals. The majority of these species are endangered or threatened. Over the last 15 years, serious concerns have been raised about sound impacts from anthropogenic noise, resulting in costly legal battles up to the level of the Supreme Court and sanctions on research in several countries. Recent research efforts have substantially increased our knowledge about some species, but we are still far from a clear understanding of hearing abilities and thus how to avoid potential sound impacts on critically endangered species
The goal of our studies has been to analyze how outer, middle, and inner ears in these animals are structured, from the gross to cytoarchitectural level, for infra (LF/IF) vs ultrasonic (HF/UF) hearing, and to determine what are the critical adaptations for detecting and localizing underwater sound as well as to produce model estimates of hearing ranges in these species. Specimens from 7 whale and dolphin species (n=26), 2 microchiropteran species (n=4), and two elephant species (n=7) were scanned as whole animals, heads, and temporal bones using spiral UHR protocols and microCT to obtain 11-100 µ isotropic voxel images of the outer, middle, and inner ears. CT analyses were conducted on Gomphothere, Mastodon, and Mammoth fossils. Cochlear radii, cochlear canal lengths, and basilar membrane dimensions were obtained by direct mapping from 2D and 3D CT reconstructions and celloidin H&E histologies.
The data show cochlear length in all groups is correlated with body mass but not with HF or LF hearing limits. Hearing ranges correlate with basilar membrane thickness and width ratios and cochlear radius of curvature in all ear types. The ears of the echolocators had significantly greater stiffness, higher basal basilar membrane ratios, and bilateral bony support up to 60% of the basilar membrane as well as foveal regions consistent with peak echolocation spectra in some species. The tighter curvatures of echolocators may reduce low frequency cochlear penetration. By contrast, elephant and baleen whale cochleae lack auxiliary stiffening agents, have similarly low apical basilar membrane ratios, and have broader cochlear radii ratios that may enhance LF propagation to the cochlear apex, suggesting common mechanical underpinnings for LF vs HF hearing. Similarities found in some fossil species suggest these frequency related adaptations evolved in parallel over similar time scales and despite media differences.
May 2017
| HRC Seminar with George Zweig | Thursday, May 4th |
George Zweig
Title: Finding the Equations of the Organ of Corti
Abstract: Measurements of basilar membrane describing functions are used to derive a nonlinear model of three dimensional cochlear mechanics. This inverse problem is solved in several steps:
1) Measurements in the nonlinear domain are extrapolated to a linear limit, establishing a basilar membrane transfer function. This process of extrapolation is aided by the discovery of unexpected regularities in the nonlinear response.
2) The transfer function is mapped to a corresponding one-dimensional partial differential equation governing cochlear traveling waves.
3) A more realistic three-dimensional model of cochlear mechanics, equivalent to the one-dimensional model in the region of the peak of the traveling wave, is established.
4) The oscillator equations for the organ of Corti found from the solution to this second inverse problem (going from one to three dimensions), is used to compute the traveling wave throughout the cochlea in the long, intermediate and short-wavelength regions. The traveling wave is predicted to have a singular secondary structure past its peak, describable in terms of a new “special function.”
5) Terms in the oscillator equations are made to depend on the oscillator’s amplitude and phase, thereby allowing the prediction of nonlinear traveling-wave responses.
6) An equation for the pressure coupling the cochlear oscillators is derived. The oscillator and pressure equations are solved for pure-tone stimuli of different amplitudes, demonstrating the nonlinear behavior of traveling waves on the basilar membrane.
Experiments to establish the validity of this approach are suggested.
September 2017
| HRC Seminar with Gin Best & Elin Roverud | September 29th |
Gin Best & Elin Roverud
Title: Investigating a visually guided hearing aid
Abstract: Understanding speech in noise continues to be the primary complaint of listeners with hearing loss. One of the few available ways to improve speech intelligibility in noise is to preferentially amplify sounds from one direction, and many sophisticated beamforming algorithms now exist that combine the signals from multiple microphones to create extremely narrow spatial tuning. In this talk we will discuss two conceptual issues with beamformers. First, the output is generally a single channel in which binaural information is lost. This has implications for locating sounds in the environment and for segregating competing sounds (e.g. in “cocktail party” scenarios). We have been exploring a hybrid strategy that combines a beamformer in the high frequencies with natural binaural signals in the low frequencies, and will describe an experiment that evaluated this strategy in different kinds of interference. The second issue is that beamformers typically emphasize one fixed direction, whereas the target of interest in many real-world situations can change location rapidly and unpredictably (e.g. in a group conversation). As a solution to this, we have developed a visually guided hearing aid, in which the user can flexibly control the acoustic look direction of the beamformer using eye gaze. We will describe two experiments that evaluate this concept using novel tasks that capture some aspects of real-world communication.
October 2017
| HRC Seminar with KC Lee | October 6th |
KC Lee, University of Washington
Title: Behavioral and neurophysiological evidence regarding the influence of oculomotor circuitry on auditory spatial tasks
Abstract: Spatial cues play an important role in segregating auditory objects in a complex acoustical scene. Spatial attention is often considered to be supramodal, e.g., crossmodal spatial cues can enhance the perception of stimuli in another modality when presented in the same location. Therefore, it is not surprising to find similarities between auditory and visual spatial attentional networks. An outstanding question, however, is how the supramodal spatial attention network functions if the listener attends instead to non-spatial acoustic features, e.g., pitch? In vision, the coupling between oculomotor circuitry and the attentional network is well studied. Are there behavioral consequences related to this tight oculomotor coupling in the context of auditory spatial tasks? We addressed these questions using three approaches. First, in a series of neuroimaging experiments using combined magneto- and electro-encephalography constrained by anatomical MRI data, we explored how different cortical regions are recruited for auditory spatial and non-spatial attention both during maintenance of attention to a single auditory stream and switching of attention between streams. Second, based on our newly developed sparse-plus-low-rank graphical approach that enables modelling of structured relationships between time series in a big data setting, we are starting work inferring functional connectivity between cortical regions to tease apart how different cortical nodes are coordinated to perform different auditory attentional tasks. Finally, we used psychophysical methods to address whether there are behavioral consequences related to this tight coupling between the oculomotor and attentional networks in the context of auditory spatial tasks.
One of the major clinical challenges facing hearing scientists is to understand why there is such great variability in performance, not predicted by standard audiological assessment, when listeners are tasked with attending to sounds in noisy, everyday settings. While our understanding of this problem, termed “hidden hearing loss,” has made great advances at cochlear and brainstem levels, our field has generally not considered how individual differences at the cortical level could also contribute to this performance variability. This is particularly pertinent to understanding why many people diagnosed with different neurodevelopmental disorders (e.g., autism spectrum disorder and fetal alcohol spectrum disorder) find it particularly challenging to communicate in noisy settings. The long-term goal of our laboratory is to understand how our auditory attentional network can be disrupted due to different neurodevelopmental deficits so we can identify potential strategies (e.g., crossmodal training) to help afflicted individuals improve social communication outcomes in their everyday lives.
November 2017
| HRC Seminar with John Rinzel | November 3rd |
John Rinzel, New York University
Title: Auditory streaming and perceptual bistability
Abstract: When experiencing an ambiguous sensory stimulus (e.g., the faces-vase image), subjects may report random alternations (time scale, seconds) between the possible interpretations. I will describe dynamical models for neuronal populations that compete for dominance through mutual inhibition, influenced by slow adaptation and noise. In highly idealized formulations network units are percept specific without direct representation of stimulus features. Our psychophysical experiments and modeling involve perception of ambiguous auditory stimuli. The models incorporate feature specificity, tonotopically organized inputs and receptive fields, so that perceptual selectivity is emergent rather than built-in. Our model addresses the effects of selective attention, distractors and deviants as well as the transient build-up phase of sound source segregation as when entering a cocktail party.
| HRC Seminar with Psyche Loui | November 10th |
Psyche Loui, Wesleyan University
Title: Creating Music by the Brain, for the Brain
Abstract: Music is a fundamentally human activity that is celebrated worldwide and from a young age, yet we are just beginning to understand and to harness its power on the human brain. I will present two lines of recent work that address the questions of how the human brain learns to create music, and how to make use of music to enhance our cognitive performance in everyday life. The study of musical creativity has been hampered by difficulties in objectively defining creative music output. Using behavioral and multimodal neuroimaging (EEG, DTI, and fMRI) studies comparing non-musicians, classically trained (non-improvising) musicians, and jazz (improvising) musicians, I will formulate a model of musical improvisation as a complex system with computational, algorithmic, and implementational levels (after Marr, 1992). Having informed a model of musical creativity, I will also present new work that harnesses the potential of music as a form of acoustic stimulation towards the goal of influencing executive function.
| HRC Seminar with Jayaganesh Swaminathan | Tuesday, November 14th |
Jayaganesh Swaminathan, Starkey Hearing Research Center
Title: Assessing the effects of reverberation on communication difficulties in hearing impaired listeners
Abstract: It has been shown that listeners with sensorineural hearing loss (SNHL) have difficulties in understanding speech in simulated reverberant environments even when wearing assistive devices such as hearing aids. However, a number of limiting factors such as poor hearing loss compensation, over simplified acoustic simulations of reverberant environments and dated statistical analysis have led to a lack of a clear understanding of the specific factors that cause hearing-impaired listeners to have difficulties in reverberant environments. In this talk, I will present results from a recently concluded study comparing normal-hearing and hearing-impaired listeners’: 1) Monaural and binaural modulation detection thresholds measured with anechoic and reverberant stimuli with different carriers and 2) Speech intelligibility in typical reverberant and noisy listening environments exploring a range of acoustic parameters, such as the level of reverberation, the type of noise and various signal-to-noise ratios. Furthermore, the effect of SNHL and reverberation on peripheral coding of acoustic features will be assessed using a physiology based model of the auditory-nerve.
December 2017
| HRC Seminar with Emily Myers | December 1st |
Emily Myers, University of Connecticut
Title: Phonetic plasticity in native and non-native speech processing
Abstract: Speech perception is subject to critical/sensitive period effects, such that the acquisition of non-native (L2) speech sounds is far more difficult in adulthood than in childhood. At the same time, adult listeners show remarkable plasticity in their speech perception system, readily adapting to new non-native and native accents and to the idiosyncratic speech details associated with different talkers. The overarching question of this work is to understand the mechanisms that underlie phonetic plasticity in native and non-native language processing, looking for commonalities among and differences between these systems. Evidence from our lab suggests that training on non-native sounds produces plastic effects in the brain regions involved in native-language perception, and that consolidation during sleep plays a large role in the degree to which training is maintained and generalizes to new talkers. Further, similar mechanisms may be at play when listeners learn to perceive non-standard tokens in the context of accented speech in one’s native language, and in perceptual adaptation to non-standard tokens. Taken together, these findings suggest that speech perception is more plastic than critical period accounts would predict and that individual variability in brain structure and sleep behavior may account for variability not only in L2 learning, but also in the way listeners approach and adapt to the speech signal in their native language.
