Date of Graduation
Doctor of Philosophy (PhD)
Valentin Dragoi, Ph.D.
Daniel Felleman, Ph.D.
William Seifert, Ph.D.
Fabrizio Gabbiani, Ph.D.
Jeremy Slater, M.D.
Psychophysical experiments in humans have demonstrated that improvements in perceptual learning tasks occur following daytime rests. The neural correlates of how rest influences subsequent sensory processing during these tasks remain unclear. One possible neural mechanism that may underlie this behavioral improvement is reactivation. Previously evoked network activity reoccurs – reactivates - in the absence of further stimulation. Reactivation was initially discovered in the hippocampus but has now been found in several brain areas including cortex. This phenomenon has been implicated as a general mechanism by which neural networks learn and store sensory information. However, whether reactivation occurs in areas relevant for perceptual learning is unknown.
To investigate how sleep affects perceptual learning at the level of single neurons and networks, an experimental paradigm was designed to simultaneously perform extracellular recordings in visual cortical area V4 along with sleep classification in monkeys. V4 is a midlevel visual area that responds to shapes, textures, and colors. Additionally, V4 is important for perceptual learning and shows significant attentional effects. In this experiment, two monkeys were trained to perform a delayed match-to-sample task before and after a 20 minute rest in a dark, quiet room. Whether monkeys exhibit the same improvements in perceptual learning previously shown in humans is unknown. Here, monkeys did improve task performance following the 20 minute rest.
Additionally, whether neural networks in V4 could reactivate was explored in a passive fixation task. A reactivation of previously evoked sequential activity was observed in V4 networks following stimulus exposure in the absence of visual stimulation. This reactivation was time-locked to when the stimulus was expected to occur after a cue, which indicated to monkeys the trial was starting. Finally, whether the delayed match-to-sample task-evoked activity was spontaneously reactivated during the 20 minute rest period was tested. No evidence to suggest that reactivation occurs during this time was observed. Considering previous reactivation results, this suggests the cue is necessary to initiate the reactivation. In summary, this work represents an investigation of the neural correlates that underlie behavioral performance improvements following daytime rest. Results can provide a better understanding of how daytime naps improve perceptual learning.
electrophysiology, visual cortex, sleep, reactivation, replay