Spatial Cues, Task Demands, and Auditory Selective Attention

Overview

Understanding speech in restaurants and other noisy environments is the most common complaint from the more than 60 million Americans with hearing impairment. Succeeding in such environments requires a listener to first segregate sound sources and then selectively attend to the source of interest. This study examines how individual auditory spatial cues (interaural time differences and interaural level differences) support spatial selective attention, the cortical processes that underlie this ability, and how these processes are modulated by task demands. Participants with normal hearing and participants who use bilateral cochlear implants will perform listening tasks while brain and physiological responses are recorded (electroencephalography, functional near-infrared spectroscopy, pupillometry, and, for normal-hearing participants only, functional magnetic resonance imaging). The findings will provide new insights into the mechanisms of spatial selective attention and help guide clinical solutions for cochlear implant users, individuals with hearing impairment more generally, and older normal-hearing adults who have difficulty understanding speech in noisy environments.

This study will examine the contributions of spatial cues (interaural level differences and interaural time differences) and task demands on spatial selective attention. Aim 1 will simultaneously collect functional near infrared spectroscopy (fNIRS), electroencephalography (EEG), pupillometry, and behavioral data in normal-hearing (NH) participants, while they perform tasks designed to probe various aspects of spatial selective attention. We will make every effort to include as many Aim 1 participants in Aim 2 as possible. Aim 2 will record functional magnetic resonance imaging (fMRI) during the same tasks. Aim 3 will essentially replicate Aim 1 in bilateral CI users. We will not record fMRI with CI users due to safety concerns. Paradigm 1. The paradigm will require listeners to hold information in working memory. We will measure effects of task demands and auditory spatial cues on cognitive and sensory responses to a single stream of stimuli. Stimuli will be a set of recorded "crash" sound effects with a broadband frequency profile, ideal for spatialization (important because a main goal of ours is to compare measures when spatialization is achieved with either interaural time differences (ITDs), which dominate spatial perception in the low-frequency region, or interaural level differences (ILDs), which dominate in the high-frequency region). Taxing working memory is an important component to the paradigm because it will allow us to influence listening effort, a critical factor in our hypotheses, and because it will ensure engagement of prefrontal cortex, which we hypothesize drives lateralization of activity in auditory cortex. Paradigm 2. The paradigm asks listeners to attend to either the Location or the Pitch of a single speech token, hold that value in memory while a diffuse noise masker is presented, and compare the remembered location or pitch to a subsequent probe token. The noise masker will ensure that listeners cannot rely on echoic memory. We will examine the effects of task demands and spatial cues on auditory cortex spatial selectivity. As with Paradigm 1 (and for the same reasons), the choice to force listeners to use working memory to perform the task was deliberate. Here, the main goal is to force listeners to have to map a perceived location to exogenous space, which is the only way to hold a location in memory. This should allow us to observe auditory cortical tuning with much more precision. Paradigm 3. This paradigm will present listeners with an ongoing, isochronous stream of words, which can be spoken by either a male or a female, and which will be spatialized to come from either the left or right. There are two tasks. The first is to attend to a given talker (male or female), and ignore location, and the second is to attend to a given location and ignore talker. On each trial, subjects must count the number of occurrences of a particular word that match the feature that they are asked to attend. The goal here is to force participants to use either space or to use another cue (voice pitch) to perform the task. Neuroimaging Approach 1. We will use EEG to record event-related potentials, which will allow us to observe changes in encoding and lateralization of sounds, which should increase as spatial cue quality increases. fNIRS will allow us to measure prefrontal cortical activity during task engagement, which has been linked to cognitive effort. Pupillometry will provide a more traditional, well-defined measure of overall effort. Behavior will provide an indication of the summative contributions of each factor. Neuroimaging Approach 2. We will use fMRI to look at stimulus-induced lateralization and spatial receptive fields in auditory cortex. We will also look for particular activity in prefrontal cortex indicating recruitment of visually-biased regions, associated with spatial processing. These regions have particular importance to our hypotheses, because they are only engaged when space is required to perform an auditory task. We propose nine experiments. Six experiments will test NH listeners, and will each leverage one of the three paradigms and one of the two neuroimaging approaches. The three additional experiments will test bilateral CI users, and we will use only neuroimaging approach 1 (fNIRS/EEG/Pupillometry), due to safety concerns regarding their medical devices and the MRI scanner.