Dissertation/Thesis Abstract

Statistical Learning of Transition Patterns between Variable Stimuli
by Dong, Mingwen, Ph.D., Rutgers The State University of New Jersey, School of Graduate Studies, 2019, 99; 13810427
Abstract (Summary)

Learning transition patterns between variable sounds is essential for vocal communications. For example, spoken speech usually consists of a series of words in a specific order. Having a variant-independent representation for a word and knowing the transition patterns between words are critical for speech perception. To investigate these questions at the neural level, we recorded extracellular neural activity from multiple sites bilaterally in the zebra finch auditory forebrain while presenting auditory stimuli in two separate experiments. In the first experiment, infrequent repetitions of a song syllable were presented after either an alternating or shuffled sequence of syllables. At all tested inter-stimulus intervals (1s, 3s, or jittered from 0.8 to 1.2s), neurons in the secondary auditory area (caudomedial nidopallium, NCM) were sensitive to the violation of transition patterns. In contrast, neurons were less sensitive to the violation of transition patterns in the primary auditory area (Field L2). These results suggest that neurons in NCM can learn transition patterns between sounds after passive exposure independent of inter-syllable intervals (at least for all tested ISIs). In the second experiment, naturally-produced variants of zebra finch songs were presented in either blocked or shuffled order. The response temporal profiles for different variants of the same zebra finch song were more similar in NCM than in L2. Furthermore, in NCM but not L2, the response temporal profiles became more similar to each other after repeated passive exposure. These results suggest that variant-independent representation emerges hierarchically in the auditory system and that passive exposure may further facilitate that representation. Together, these two experiments provide insights into how the zebra finch auditory system can form variant-independent representations of complex sounds and learn the transition patterns between those sounds. Because similar neural mechanisms may serve the statistical learning and perceptual invariance capacities of the human auditory system, this approach may help us understand the neural basis of speech perception and ultimately contribute to treatments for certain auditory processing disorders.

Indexing (document details)
Advisor: Vicario, David S
Commitee: Bieszczad, Kasia M, McGann, John P, Otchy, Timothy M, Vicario, David S
School: Rutgers The State University of New Jersey, School of Graduate Studies
Department: Psychology
School Location: United States -- New Jersey
Source: DAI-B 81/3(E), Dissertation Abstracts International
Source Type: DISSERTATION
Subjects: Neurosciences, Cognitive psychology
Keywords: Oddball, Songbird, Statistical learning, Surprise, Transition pattern
Publication Number: 13810427
ISBN: 9781088326206
Copyright © 2019 ProQuest LLC. All rights reserved. Terms and Conditions Privacy Policy Cookie Policy
ProQuest