Caltech Library System

About | Browse | Search | Caltech Student Instructions

Type of Document	Dissertation
Author	Bray, Signe Lauren
URN	etd-07052008-163103
Persistent URL	http://resolver.caltech.edu/CaltechETD:etd-07052008-163103
Title	Neural mechanisms underlying the influence of associative learning on valuation and decision-making in humans
Degree	PhD
Option	Computation and Neural Systems
Advisory Committee	Advisor Name Title Ralph Adolphs Committee Chair Antonio Rangel Committee Member John O'Doherty Committee Member Richard Andersen Committee Member Shinsuke Shimojo Committee Member
Keywords	fMRI associative learning
Date of Defense	2008-07-02
Availability	mixed
Abstract Reward is a powerful modulator of behavior. Animals and humans are endowed with the ability to learn to associate events and actions with reinforcing stimuli, and flexibly adapt their behavior. The experiments described in this thesis use functional magnetic resonance imaging (fMRI) to study the neural mechanisms of reward learning in humans, the neural substrates by which reward associations influence behavior, and the neural plasticity that can be induced by provision of reward. Attractive faces have been shown to be a form of visual reward, but their influence on behavior has yet to be characterized. We show that reward prediction errors in the nucleus accumbens are engaged when subjects learn associations between neutral cues and attractive faces, as has been shown with other reinforcers such as juice and money. This learning increases the subjective value of cues associated with attractive faces. Animal studies have shown that Pavlovian cues can influence response vigor and decision-making. We present the first investigation into the neural mechanisms by which Pavlovian cues influence human decision-making. We find that activity in the ventral striatum differentiates between decisions to act in a manner compatible or incompatible with a concurrently presented Pavlovian cue. In the next section we apply associative learning techniques to directly instrumentally condition neural activity, using reward feedback derived from fMRI images processed and analyzed in real time. This technique presents an alternative to standard bio/neurofeedback approaches and may prove useful in many clinical and research applications. We demonstrate that this method can be used to probe the causal influence of regional brain activity; specifically we test the impact of medial orbitofrontal cortex (mOFC) activity on affective judgments. Subjects learn to elevate mOFC activity on cue and elevated mOFC activity increases the propensity to make a positive subjective valuation judgment. Taken together these studies advance our understanding of the functional contributions of ventral striatum and orbitofrontal cortex in influencing decision-making and valuation, and illustrate the utility of applying associative learning techniques in combination with real-time fMRI in order to evaluate the causal contribution of specific brain regions toward particular cognitive functions.
Files	Filename       Size       Approximate Download Time (Hours:Minutes:Seconds)     28.8 Modem   56K Modem   ISDN (64 Kb)   ISDN (128 Kb)   Higher-speed Access  thesis_fixdup.bib 80.65 Kb 00:00:22 00:00:11 00:00:10 00:00:05 < 00:00:01 thesis_sb_jul508.pdf 3.27 Mb 00:15:07 00:07:46 00:06:48 00:03:24 00:00:17 indicates that a file or directory is accessible from the campus network only and must not be distributed to non-campus persons.