Abstract
Perceptual learning is a persistent and specific improvement in the performance of a perceptual discrimination. Previous research found that perceptual learning of a texture discrimination task is sleep-dependent (Karni et al., 1994) and that repeated, within-day testing leads to perceptual deterioration that is specific to the trained condition (Mednick et al., 2005). Daytime naps mitigate this deterioration (Mednick et al., 2002) and can even lead to improvement in task performance within the same day (Mednick et al., 2003). We studied the effects of daytime naps on perceptual learning of a motion direction discrimination task (Rokem and Silver, 2010). Following training on this task, subjects were classified into one of four groups: 50–90 min naps with or without rapid eye movement (REM) sleep, measured with polysomnography (PSG), Quiet Wake (75 min with eyes closed, no sleep, monitored with PSG), and Active Wake (subjects who continued their day as usual). Similar to previous findings of perceptual deterioration in the trained condition, we found decreased performance in the trained motion direction without sleep (all subjects, prior to the nap). In addition, subjects who napped (REM + non-REM) exhibited increased perceptual learning in the trained direction of motion, after the nap, relative to non-nappers (active wake + quiet wake). No such effect of napping was found for learning in an untrained direction of motion. Furthermore, among REM nappers, the number of minutes spent in REM sleep was significantly correlated with the specificity of perceptual learning (defined as the difference between the amount of learning in the trained direction and in the untrained direction). We conclude that sleep plays an important role in the consolidation of learning of motion perception and that REM sleep promotes stimulus specificity of perceptual learning.