Barbara G. Shinn-Cunningham - US grants

DESCRIPTION (provided by applicant): The goal of this project is to characterize and understand how spatial auditory perception is altered by past experience. Experiments will test the hypothesis that spatial auditory processing is hierarchical and that experience influences spatial perception through different mechanisms at different processing stages. Specifically, experiments will test whether 1) explicit training rapidly alters how spatial percepts are mapped to perceived locations in exocentric space; whereas 2) implicit learning of room acoustics alters how spatial cues are integrated to form spatial percepts, leading to improvements in localization reliability with experience in a room; and 3) early stages of the pathway do not change, but include dynamic, nonlinear effects that alter what information is available to later processing stages. These hypotheses will be addressed through behavioral experiments that measure spatial localization accuracy and variability, acoustic measurements that characterize how spatial acoustic cues are affected by different acoustical environments, and computational models that explore what processing structures account for experimental results. This work is critical for identifying how the normal human listener adapts and calibrates spatial perception in everyday settings in order to maintain accurate spatial perception in changing acoustic environments. Results will help to elucidate how past experience influences spatial auditory processing, changes spatial auditory perception, and affects spatial resolution and localization accuracy. By studying perceptual phenomena that reflect changes at different stages of processing and affect behavior over a range of time scales, this project will provide a coherent picture of how the normal auditory system keeps spatial auditory percepts in registry with percepts from other modalities.

The Center of Excellence for Learning in Education, Science, and Technology (CELEST) brings together leading scientists, educators, and technologists from Boston University, Brandeis University, Massachusetts Institute of Technology, and the University of Pennsylvania to study real-time autonomous learning systems by integrating experimental and computational brain science, biologically inspired technology, and classroom innovation. Contributing scientists are drawn from four Boston University Departments and the Center for Adaptive Systems, the Center for Memory and Brain, the Science and Mathematics Education Center, the Hearing Research Center, and the Center for Polymer Studies; the Brandeis University Department of Psychology and the Volen Center for Complex Systems; the MIT Department of Brain and Cognitive Sciences, the Picower Center for Learning and Memory, and the Harvard/MIT Speech and Hearing Bioscience and Technology Program; and the University of Pennsylvania Department of Psychology.
Intellectual Merit and Creative Concepts: CELEST brings together educators, scientists, and technologists to carry out four types of mutually reinforcing and integrated activities: (1) quantitative behavioral and brain modeling of both normal and abnormal learning processes during perception, cognition, emotion, and action; (2) interdisciplinary cognitive and neuroscience experiments to probe these processes and to test model predictions; (3) development of algorithms, based on biological learning models, for incremental fast learning about complex and rapidly changing environments in large-scale engineering and technological applications that are important in many areas of society; and (4) integration of research and education through contributions to educational technology, curriculum development, and early career recruitment of underrepresented communities into scientific practice. These goals are achieved through interactions among eight main Thrusts in:
Learning in visual perception and recognition: laminar cortical dynamics of adaptive behavior;
Learning in audition, speech, and language; learning in cognitive-emotional interactions and planned sequential behaviors;
Learning and episodic memory: encoding and retrieval;
Learning in concept formation and rule discovery;
Learning in attentive recognition and neuromorphic technology;
Educational technology, curriculum development, and outreach; and
Diversity outreach.
Broader Impact: CELEST will foster interdisciplinary collaborations and training across all its units: frequent seminars, workshops, colloquia, conferences, and publications; integration of research and education by translating basic science results into interdisciplinary curriculum development; and web-based and hands-on training for teachers and students, including classroom activities with a national and international impact. CELEST will hereby provide world-class expertise towards advancing key SLC program goals; namely, the psychological and pedagogical aspects of learning, the biological basis of learning, machine learning, learning technologies, and mathematical analyses and modeling of them all.

[unreadable] DESCRIPTION (provided by applicant): The current project explores dynamic processes in spatial auditory perception. Specifically, it studies how representations of auditory space adapt to the auditory context defined by the preceding stimulation and by the listener's behavioral task. A series of behavioral experiments and computational modeling studies is proposed. The first tested hypothesis is that the contextual plasticity is influenced not only by the bottom-up factors, like the spatial distribution of the stimuli, but also by the top-down factors, for example, by the listener's concentration on the expected target locations and away from the distractor locations. This will be tested by comparing the plasticity observed when the listener is performing a task that requires specific distribution of spatial attention to plasticity observed when attention does not need to be focused. The second hypothesis is that the contextual plasticity is influenced by the spatial and temporal distribution of the stimuli. This will be tested by measuring plasticity observed in auditory scenes with various spatial and temporal distributions of distractor and target stimuli. The third hypothesis is that the anatomical locus of the contextual plasticity is at a central level of the auditory pathway where the underlying binaural cues are integrated into spatial percepts. This will be tested by evaluating the extend to which contextual plasticity induced by interaural temporal cues generalizes to interaural intensity cues for auditory localization, and vice versa. The results of these experiments will be evaluated in the context of the hierarchical model of auditory spatial perception proposed in the parent grant. Previous studies found that short-term plasticity in spatial auditory representations can be induced by sustained presentation of constant stimuli or by visual feedback to which the listeners align the auditory percepts. Other than that, spatial auditory perception was thought of as being static. The existence of contextual plasticity explored in this project is important because it suggests that auditory spatial representation is dynamic and that it changes continuously depending on the listener's task and on the auditory scene. Understanding how normal-hearing listeners adapt to the auditory context is critical, e.g., because similar plastic changes might take place in the central auditory processing of listeners with impaired auditory periphery. And designing prosthetic devices that predict and adapt to this neural plasticity might significantly improve the performance of such devices. This research will be done primarily in Slovakia at the Technical University of Kosice in collaboration with Norbert Kopco, as an extension of NIH grant #5R01 DC005778-03. [unreadable] [unreadable] [unreadable]

[unreadable] DESCRIPTION: Hearing-impaired listeners and cochlear implant users have difficulty selectively attending to a sound source in environments with multiple, competing sound sources. In contrast, normal-hearing listeners are able to selectively attend to a talker of interest in a sea of competing sources, shifting attention as the need arises. The long-term objective of this project is to develop strategies and devices to help impaired listeners communicate effectively in everyday social settings that require selective attention. The current proposal is built on the hypothesis that some of the difficulty that impaired listeners have in complex environments stems from difficulties in properly forming auditory "objects" (i.e., grouping sound coming from one physical source into a single perceptual entity). Because interactions between auditory grouping and auditory perception are poorly understood, a necessary first step towards our long-term objective is to understand how the grouping of sound into objects affects perception in normal-hearing listeners. We will perform psychophysical experiments in normal- hearing listeners that test how grouping influences auditory perception and use computational models to understand the processes that normally affect how listeners attend to one source in a mixture of competing sound sources. The specific aims of this project explore how perceptual grouping influences perceptual sensitivity to basic psychophysical sound attributes (Aims 1 and 2) as well as the ability to understand speech when there is a competing, similar sound source (Aim 3). Aim 1 explores how sensitivity to spatial cues and the intensity of one tone can be degraded when that tone is perceptually integrated into an object containing other sounds. Aim 2 explores how the ability to extract the spectro-temporal pattern defined by multiple simultaneous or sequential target tones (e.g., spectral profile, rhythm, pitch contour) can be degraded when competing sounds interfere with the perceptual grouping of the target tones into a perceptual object. Aim 3 measures how weakened grouping cues (like those commonly experienced with hearing impairment or through a cochlear implant) interfere with speech intelligibility in the presence of a competing sound object. Together, the proposed studies test a unified conceptual model of why perceptual grouping directly influences the ability to communicate in the complex social settings, from heated business meetings to lively dinner parties. PUBLIC HEALTH RELEVANCE One of the most common complaints of hearing aid and cochlear implant users is that they cannot communicate effectively in everyday settings where there are multiple, competing sound sources. This problem often causes social isolation, as listeners opt out of trying to participate in complex settings rather than facing frustration and failure. Understanding the processes that allow normal- hearing listeners to cope with complex acoustic environments is a critical and necessary step towards developing signal-processing schemes and devices to ameliorate the difficulties facing impaired listeners. [unreadable] [unreadable] [unreadable]

CELEST seeks to understand the fundamental processes that underlie human learning by studying dynamic interactions within and among brain regions. Interdisciplinary research teams study how the brain learns to (1) plan: to make decisions for appropriate actions based on assessment of risks and potential rewards in a given situation, (2) explore: to perform planned actions to move about familiar and unfamiliar environments, (3) communicate: to use noisy and incomplete sensory information to interact effectively with other agents and objects in the world, and (4) remember: to encode and guide retrieval of information to achieve goals. CELEST is a multi-faceted collaboration that focuses the efforts of scientific and educational teams led by 15 senior scientists at four Boston-area universities. CELEST combines undergraduate and graduate training in interdisciplinary research that combines experimental cognitive neuroscience with quantitative behavioral and brain modeling of normal and abnormal learning during perception, cognition, emotion, and action.

Broader impacts: CELEST transfers the results of basic research on learning to undergraduate and graduate courses. This is achieved through its ongoing development of course materials for the new undergraduate neuroscience major at Boston University, and through electronic dissemination on the CELEST web site. Outreach to the undergraduate neuroscience community also occurs by means of a one-day CELEST workshop and related workbook about the cognitive basis of successful learning strategies. A number of CELEST programs are targeted at increasing opportunities for groups underrepresented in science to participate in its innovative curriculum and research initiatives. These include graduate fellowships, summer internships for faculty from minority-serving institutions, a ten-week summer program for undergraduates from underrepresented groups to work in CELEST faculty labs, and a week-long summer workshop to introduce undergraduates to the interplay of modeling and experimental techniques in cognitive neuroscience. Center added value: By bringing together distinct scientific communities that traditionally employ different practices and techniques, CELEST interdisciplinary science is changing the way we understand how the brain learns, and how different parts of the brain interact with each other during learning. Through collaboration with industrial partners, including the development and transfer of large-scale neuromorphic engineering and technological algorithms to industry and government laboratories, CELEST facilitates research for practical applications that cannot be supported by conventional single-investigator grants. CELEST faculty, postdocs, and students are playing increasingly important roles in communicating with non-specialists through many activities including blogs, workshops, and presentations to secondary school audiences. The integration of CELEST research and education is accomplished through the development of innovative curriculum materials based upon mathematical and computational models of mind and brain. through electronic and personal presentations to a variety of audiences, and through sponsorship of scientific conferences and workshops.

This grant will support a two-and-a-half-day long workshop on Computational Audition, bringing together top scientists from a broad range of disciplines. Computational audition, the study of how information can be derived from sound in both biological organisms and in machines, is an emerging field. The topic is broad, encompassing a diverse group of scientists including neuroscientists, psychologists, psychophysicists, speech scientists, computer scientists, and engineers. The field of computational audition has the potential to be a model of interdisciplinary research with a great deal of intrinsic intellectual interest. It is primed to become a hotbed of scientific growth, and this meeting will help the field evolve toward that goal.

Broader Impacts
Computational audition is important both for developing new technological applications as well as for finding new clinical treatments (because understanding the basis of normal hearing will help treat hearing impairment). And yet by comparison to many other related fields, its potential is underexplored. For instance, research on human and machine perception has been dominated to a large extent by vision, with hearing neglected by comparison. This workshop will nurture the development of interdisciplinary research in computational audition that will capitalize on this potential.

One of the key features of the brain is how quickly and naturally it learns from experience -- not only factual information, but also abstract relationships and associations. Fire and flame burn; a sound like "mama" typically conjures up concepts of mother, comfort, and love. While recent advance in neuroscience have helped to uncover the basic brain processes that support everyday learning, this knowledge has not yet realized its potential to help people facing various clinical challenges. The goal of this project is to bring together neuroscientists who understand the neuroscience of learning, rehabilitation specialists who understand what kinds of problems individuals face, and computer scientists who can help create engaging training regiments that make it painless and even fun to undertake the often repetitious training required to support learning. The assembled team will support exchanges of pre- and post-doctoral trainees across different laboratories to jump-start collaborations, seed projects that support interdisciplinary research into engaging training paradigms, a web portal to disseminate results, and an annual scientific conference that brings together the diverse expertise necessary to create fun and interesting neural training programs. Immersive, active training programs can have beneficial effects on the brain, and can be harnessed to study the process of learning in motivated, engaged participants. Yet realizing this potential requires bringing together traditionally distinct research communities. Collaboration is needed among neuroscientists (node 1) studying the science of learning; rehabilitation researchers (node 2) working with individuals who benefit from brain training; and computer scientists (node 3) who know how to build computer software that is fun to use. The Engaging Learning Network (ELN) will develop a collaborative web encompassing these three nodes of expertise. ELN will extend and apply the neuroscience of learning through training programs that enchant a diversity of users while enhancing their sensory, motor, and/or cognitive function.

The network comprises 24 investigators from 15 institutions: 10 neuroscientists, 9 rehabilitation researchers, and 5 computer scientists. Collaborations will be built through a scientific meeting, seed projects, graduate student exchanges, and a web portal. The annual scientific meeting will engage members of all ELN laboratories as well as others from outside the network to promote multi-disciplinary research on the art of designing captivating training programs that enable us to study neural plasticity using programs that enhance human function. Results will be disseminated through scientific meetings and the web. Once vetted through peer review, many participants will make sensori-motor training applications commercially available; these synergistic efforts will enhance the real-world impact of the network activities and leverage the basic science supported by the project.