Kamal Sen - US grants

The goal of this proposal is to investigate the auditory processing of natural sounds. This problem will be investigated in the auditory forebrain of awake, behaving songbirds. The statistical structure of natural sounds will be analysed using a variety of quantitative techniques, and synthetic sounds that preserve or alter specific aspects of natural sounds will be generated. Electrophysiological recordings in a chronic preparation will be obtained, using mufti-electrode arrays, from a population of neurons in successive stages of the auditory pathway. New theoretical methods for the analysis of the neural coding of natural sounds will be applied to quantitatively describe the features of sounds to which auditory forebrain neurons are sensitive and information theory will be applied to quantify the amount of information conveyed by the neural responses. Processing capabilities of the auditory circuits that are innate as well as those that depend on learning and development will be investigated by comparing the neural responses in the juvenile and the adult auditory systems. Such circuitry is analogous in a variety of ways to the innate neural networks that must exist in human infants and endow them with innate responsiveness to speech sounds of all human languages at birth. Thus, the results of this investigation will be relevant to disorders involving deficits in speech acquisition and learning.

[unreadable] DESCRIPTION (provided by applicant): Auditory cortex plays an important role in the processing of complex sounds e.g., vocal communication sounds and human speech. Cortical organization is altered by deafness and cortical plasticity is thought to play a critical role in recovery of speech perception e.g., in children with cochlear implants. Our long-term goal is to understand the neural basis for cortical processing of complex sounds and associated plasticity. A fundamental problem faced by animals and humans is to discriminate between complex sounds e.g., species-specific sound. However, cortical neural mechanisms underlying such discrimination are poorly understood. How reliably do cortical neurons discriminate between communication sounds? How does neural discrimination change during development, depend on early acoustic experience, and recover from early auditory deprivation. In this proposal we begin to investigate these questions in songbirds, a model system that offers unique advantages, with particular relevance to human speech. We will probe neural discrimination in field L, the avian analogue of primary auditory cortex, which is thought to play an important role in the perception of vocal communication sounds. This proposal will focus on adult male zebra finches. We will take an integrative approach combining experiments, theoretical analysis and computational modeling. In Aim 1, we will record neural responses to conspecific songs. We will then use theoretical methods to quantify how reliably these different sounds can be discriminated, based on single neuron responses, analyzing how discrimination evolves over time, and how it depends on the temporal precision of neural responses, We will test if spike timing improves discrimination accuracy. In Aim 2, we will characterize the receptive field (RF), i.e., the features of sound to which a neuron responds, identify critical RF parameters that may improve discrimination and assess the contribution of non-linear response; components to neural discrimination. We will test whether the RF is predictive of discrimination. In the long term, we will examine processing of vocal communication sounds in young birds, the effects of rearing birds under abnormal acoustic conditions e.g., isolation, and the recovery from auditory deprivation. Ultimately, understanding the capacity, as well as limits, of cortical plasticity in processing vocal communication sounds, may help guide recovery in humans with deficits in speech perception. [unreadable] [unreadable] [unreadable]

In everyday social situations, normal hearing humans are able to listen to a speaker in the midst of other people talking and other sound sources. This is an example of a general problem termed auditory scene analysis. The understanding of auditory scene analysis remains a challenging problem in a diverse range of fields such as neuroscience, computer science, speech recognition and engineering, after more than 50 years of research. Although a difficult problem for machines and hearing impaired listeners, humans with normal hearing solve it with relative ease. This suggests the existence of a solution to this problem in the brain, but this solution remains unknown. This project will investigate how the brain solves this problem. By revealing circuits in the brain that contribute to the solution, this project will ultimately improve quality of life through applications in medical devices, e.g., hearing aids and cochlear implants; benefit society through applications in technology, e.g., applications for speech recognition; and create an educational platform to train students to integrate knowledge from a variety of disciplines to address challenging and important societal problems.

The spatial location of different sound sources is an important component of auditory scene analysis. The auditory cortex, with its unique spatial sound processing ability, is thought to play an important role in auditory scene analysis, although the underlying neural network mechanisms remain largely unknown. This project will investigate cortical circuits for auditory scene analysis in the primary auditory cortex of the mouse, employing powerful optogenetic tools to investigate both bottom-up and top-down mechanisms. First, the investigators will examine the influence of behavioral states on cortical spatial representations of sound mixtures across different cortical layers and test the hypotheses that such spatial representations vary across cortical layers and behavioral states of the animal. Second, the investigators will examine the causal role of parvalbumin positive (PV) inhibitory interneurons versus somatostatin positive (SOM) inhibitory interneurons in the primary auditory cortex, using optogenetic manipulations, and test the hypothesis that PV interneurons are critical in mediating bottom-up signaling, whereas SOM interneurons are selectively engaged during active behavioral states. Finally, the investigators will construct a computational model of the primary auditory cortex including excitatory, PV and SOM neurons to explain the experimental results and make predictions for future experiments.

This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.

The ability of our auditory systems to recognize target sounds in a mixture of other sounds is fundamental to normal healthy function and communication. For example, during the course of a normal day we must communicate with a conversation partner in the presence of other sounds, e.g., other people talking, music, sound of cars etc. Like humans, many animals are capable of listening to a single sound source in a mixture of sources. Thus, neural circuits for solving the CPP also likely exist in animals. This proposal, will investigate how the auditory cortex contributes to solving this problem by unraveling cortical circuitry that underlies the processing of complex sound mixtures, using powerful experimental tools available in mice. By providing new insights into cortical mechanisms that help solve this problem in the normal brain, this proposal may impact the development of novel therapeutic strategies for the hearing impaired, who have great difficulty solving this problem, and improve hearing assistive devices, e.g., hearing aids and cochlear implants.