Abstract
Little is known about how neurochemical systems are involved in human learning despite clear evidence that these systems are crucial to the plasticity underlying learning in animals. For example, norepinephrine (NE) is thought to act as a signal that "tells" the brain when to learn. Here we explore the hypothesis that NE is involved in a fast-Task Irrelevant Perceptual Learning paradigm using pupil dynamics as a surrogate measure of neurochemical activity. This hypothesis is motivated by evidence that pupil dynamics in monkeys are coupled with locus coeruleus activity (LC; the center of NE release). In this paradigm, participants conduct dual target and image recognition tasks each trial. In the target-recognition task, participants are presented with a stream of images temporally paired with an alphanumeric Target (e.g., a number) or Distractor (e.g., letters). Participants report the identity of the Target from a stream of Distractors and then report which of two images was in the image-stream. Each Target or Distractor paired image is presented for 133ms with an ISI of 1000ms, and pupil metrics are continuously recorded every 10ms. Participants showed increased recognition accuracy when the tested image was paired with a Target in comparison to tested images paired with Distractors. Pupil dynamics also indicated that pupil size changed more after Target processing compared to Distractor processing. Furthermore this effect was specific to subjects showing the improvement in memory processing for target-paired stimuli. A separate experiment utilized an unexpected sound to induce pupil size changes, confirming the relationship between pupil-size changes and memorization rates. Together, these results demonstrate a relationship between pupil size changes and task-irrelevant learning. While further research is required to demonstrate a casual link between NE activity and pupil dynamics, these results are consistent with the hypothesis that NE plays a role task-irrelevant perceptual learning.
Meeting abstract presented at VSS 2014