Most real-world sounds differ from each other in the pattern with which sound level is distributed across audio frequency. Accurate perception of this spectral envelope aids the ability to distinguish between sounds. However, limiting this ability, the internal representation of the spectral envelope is often degraded in normal-hearing listeners in difficult listening environments, and in individuals with hearing disorders. Of interest here is the possibility that the capacity to detect peaks and valleys in a spectral envelope can be improved through practice. The long-term goal of this application is to develop perceptual-training programs that improve the perception of real-world spectral envelopes, based on an understanding of the neural circuitry underlying spectral-envelope perception. The current objective is to determine the extent to which multiple-hour training improves the ability to detect a simple sinusoidal distribution of spectral peaks and valleys (a single spectral-modulation frequency), both in sounds in which that spectral envelope is constant and in those in which it changes over time (spectro-temporal envelope), and to use the pattern with which this learning generalizes to untrained stimuli to make inferences about the neural circuitry modified by training.
The specific aims are designed to use behavioral techniques to infer whether training on a single modulation detection condition (spectral or spectro-temporal) modifies neural circuitry (1) that is tuned to the trained spectral modulation frequency [Aim 1], range of audio frequencies [Aim 2], and rate and direction of spectro-temporal modulation [Aim 3], and (2) that subserves the perception of speech in noise [Aim 4]. To pursue these aims, in each of three experiments, normal-hearing adults will be trained to detect spectral or spectro-temporal modulation in a single condition and tested on that condition and several untrained ones before and after the multiple-hour training. The underlying assumption is that learning on the trained condition will generalize to an untrained condition if and only if the neural circuitry modified during training also influences performance on the untrained condition. Relevance: The knowledge gained through the proposed studies will be a first step toward developing optimized auditory-training programs that improve the ability to detect peaks and valleys of sound level across frequency. Because this ability, while fundamental to the perception of real-world sounds, is often compromised in normal-hearing listeners in noisy environments, and in individuals with hearing disorders, such training programs have the potential to improve public health. ? ? ? ?
| Andéol, Guillaume; Macpherson, Ewan A; Sabin, Andrew T (2013) Sound localization in noise and sensitivity to spectral shape. Hear Res 304:20-7 |
| Sabin, Andrew T; Clark, Cynthia A; Eddins, David A et al. (2013) Different patterns of perceptual learning on spectral modulation detection between older hearing-impaired and younger normal-hearing adults. J Assoc Res Otolaryngol 14:283-94 |
| Sabin, Andrew T; Eddins, David A; Wright, Beverly A (2012) Perceptual learning of auditory spectral modulation detection. Exp Brain Res 218:567-77 |
| Sabin, Andrew T; Eddins, David A; Wright, Beverly A (2012) Perceptual learning evidence for tuning to spectrotemporal modulation in the human auditory system. J Neurosci 32:6542-9 |
| Sabin, Andrew T; Hardies, Lauren; Marrone, Nicole et al. (2011) Weighting function-based mapping of descriptors to frequency-gain curves in listeners with hearing loss. Ear Hear 32:399-409 |
