Complex pitch perception in complex environments

Pitch plays a fundamental role in auditory and speech perception. In music it defines melody and harmony; in speech it carries prosodic and (in tone languages) lexical information; and in everyday complex acoustic environments it plays a critical role in helping us to segregate and hear out important sounds against a background of other sounds. Despite its importance, fundamental questions surrounding how pitch is perceived and neurally encoded remain unanswered. There is, for instance, no consensus on how and why pitch accuracy degrades with hearing loss and/or age. Similarly, despite cochlear implants’ many successes, pitch remains very poorly conveyed for reasons that remain a matter of debate. This project combines behavioral studies in normal- hearing and hearing-impaired adults across the lifespan (Aim 1), behavioral and EEG studies in infants (Aim 2), and high-resolution neuroimaging (fMRI) studies in adults (Aim 3) with computational modeling to answer a cohesive set of fundamental questions on how pitch is perceived and represented. These questions include whether peripheral or more central (experienced-based) factors lead to the dominance of low-numbered harmonics in everyday pitch perception (Aims 1 and 2), and whether the interactions between the perceptual dimensions of pitch (based on fundamental frequency) and brightness (based on spectral centroid) follow a developmental trajectory that is also reflected in their cortical representations (Aims 2 and 3). The first two experiments of Aim 1 (Exp. 1A/B) use individual differences to determine the effects of age and hearing loss on the accuracy of pitch perception. They test the hypotheses that the sharp transition between good and poor pitch discrimination as lower harmonics are progressively removed is due to the limits of auditory filtering, as influenced by cochlear hearing loss, and that overall pitch fidelity across all conditions is degraded by age, as expected from age-related cochlear synaptopathy. Aim 1’s third experiment (Exp. 1C) utilizes recent signal- processing developments to test the role of harmonicity of either the target or interfering speech in the perceptual segregation of competing talkers. Aim 2 uses a developmental approach to test whether properties that are currently believed to be innate, including interference between pitch and brightness (Exp. 2A/B), and poor accuracy at high harmonic numbers and/or frequencies (Exp. 2C/D), are in fact centrally mediated developmental traits that emerge after extensive exposure to natural sounds. Finally, Aim 3 uses ultra-high-field (7T) fMRI to determine whether the perceptual interference observed in adults between pitch and brightness is reflected by interdependence of cortical tuning (Exp. 3A); whether cortical mapping of pitch and frequency is affected by the context in which the tones are presented (Exp. 3B); and whether cortical correlates of pitch interval size and direction can be observed (Exp. 3C). By improving our fundamental understanding of the neural coding and perception of pitch, we will be better placed to improve its representation in people with hearing loss and cochlear implants.