LDR | | 00000nam u2200205 4500 |
001 | | 000000435545 |
005 | | 20200228101154 |
008 | | 200131s2019 ||||||||||||||||| ||eng d |
020 | |
▼a 9781392585245 |
035 | |
▼a (MiAaPQ)AAI13810363 |
040 | |
▼a MiAaPQ
▼c MiAaPQ
▼d 247004 |
082 | 0 |
▼a 150 |
100 | 1 |
▼a Sha, Long. |
245 | 10 |
▼a Psychophysical and Neural Mechanisms of Perceptual Learning. |
260 | |
▼a [S.l.]:
▼b New York University.,
▼c 2019. |
260 | 1 |
▼a Ann Arbor:
▼b ProQuest Dissertations & Theses,
▼c 2019. |
300 | |
▼a 209 p. |
500 | |
▼a Source: Dissertations Abstracts International, Volume: 81-05, Section: B. |
500 | |
▼a Advisor: Kiani, Roozbeh. |
502 | 1 |
▼a Thesis (Ph.D.)--New York University, 2019. |
506 | |
▼a This item must not be sold to any third party vendors. |
506 | |
▼a This item must not be added to any third party search indexes. |
520 | |
▼a In a jungle, the faint sound of a snapping twig could mean a stalking tiger. On a highway, a minor swerve of a nearby car could mean an impending accident. Our brain must learn to use subtle sensory information to make precise and prompt responses. How the brain achieves high level of perceptual expertise through experience, a process known as perceptual learning, remains poorly understood. Our current understanding builds on behavior of learning simple perceptual stimulus, and measurements from single neurons in sensory regions during learning. However, neurophysiological evidence suggests that activity in decision-making regions, rather than sensory regions, show changes closely related to changes in behavior due to learning. In addition, recent theoretical studies suggest that changes in neural populations could have much larger effects on brain functions than changes of individual neural responses. It is, therefore, critical to study the mechanisms of perceptual learning from the perspective of neural population activity in decision-making regions.I pursue this goal in my thesis by developing an experimental and theoretical approach to study how neural populations in decision-making regions underlie perceptual learning. My research creates a framework for comprehensively measuring perceptual learning, developing computational models, and discovering neural mechanisms that implement these computations.Goal 1: I have developed a novel task in which humans and non-human primates learn to categorize unfamiliar visual shapes composed of multiple features. Through experience, subjects become better at discerning small feature variations that determine category membership of the shapes. My task design enables me to identify different processes that contribute to learning, and define the time course and interactions of those processes. For example, by measuring how a combination of small feature variations influences subjects' choices, I can probe how subjects weigh the features to define categories and where they believe the true category boundary is at each point during learning. My results reveal three key processes: improved sensitivity to sensory information, improved readout of these representations for categorization, and improved decision strategies for utilizing the sensory information. Because past studies focused on these processes in isolation, their relative contribution to learning is currently unknown. Similarly, the time course of changes in these processes remains undefined. Further, it is unclear how these processes are implemented by neural populations in the brain. My task design addresses these questions and my computational models provide a normative framework for studying them. These models also predict how neural populations in sensory and motor cortices should adapt during learning. I test these predictions in my second goal.Goal 2: To characterize the neural mechanisms of learning at the level of neural populations, I have trained two macaque monkeys to perform my task, while I record simultaneously from neural population responses in decision-making cortex (frontal eye fields) and two visual cortices (V4 and posterior inferior temporal cortex, not included in this thesis). I have developed statistical analyses to quantify task information encoded by these neural populations and their correspondence to behavioral changes during learning. My results indicate widespread changes in cortex that implement the learning mechanisms mentioned above. Neurons in decision-making cortex improve readout of the decision-making process to optimize behavior based on past experience.Broader Significance: Learning is the essence of our adaptive behavior. It plays a key role in our ability to improve based on past experience in order to adjust to changes in our environment. My study aims to clarify the principles that implement perceptual learning in large biological neural networks. Knowledge from my study has widespread implications both for understanding disorders of learning and for devising new teaching methods for faster and more effective learning.Dissertation Chapters: My dissertation consists of a general introduction, 4 thesis chapters, and a general conclusion. In Chapter 2, I outlined the behavioral paradigm and human psychophysics results. In Chapter 3, I described two monkeys' performance at learning a similar task introduced in Chapter 2. In Chapter 4, I showed neural results from monkey frontal eye fields during perceptual learning. The ongoing work on IT and V4 recording would not be included in the dissertation. In Chapter 5, I included details in monkey training, neural recording and spike sorting crucial for understanding the details in my results and for replicating the study. |
590 | |
▼a School code: 0146. |
650 | 4 |
▼a Neurosciences. |
650 | 4 |
▼a Psychology. |
690 | |
▼a 0317 |
690 | |
▼a 0621 |
710 | 20 |
▼a New York University.
▼b Center for Neural Science. |
773 | 0 |
▼t Dissertations Abstracts International
▼g 81-05B. |
773 | |
▼t Dissertation Abstract International |
790 | |
▼a 0146 |
791 | |
▼a Ph.D. |
792 | |
▼a 2019 |
793 | |
▼a English |
856 | 40 |
▼u http://www.riss.kr/pdu/ddodLink.do?id=T15490645
▼n KERIS
▼z 이 자료의 원문은 한국교육학술정보원에서 제공합니다. |
980 | |
▼a 202002
▼f 2020 |
990 | |
▼a ***1008102 |
991 | |
▼a E-BOOK |