Abstract
The adult human brain shows remarkable plasticity following perceptual learning, resulting in improved visual sensitivity. Such improvements in visual perception commonly require extensive practice, predominantly mediated by practice-dependent plasticity in early visual areas. However, the common practice-makes-perfect dogma as a unitary form of procedural learning has recently been challenged. Adapting memory reactivation-reconsolidation frameworks, we showed that brief reactivations of an encoded visual-skill memory are sufficient to improve human perceptual thresholds (Amar-Halpert et al., 2017). Here, we aimed to reveal the underlying mechanisms of reactivation-induced perceptual learning. For this purpose, 40 participants performed a standard texture discrimination task (TDT) (Karni and Sagi, 1991). The memory was encoded and consolidated on a Day1 standard session (252 trials), during which the discrimination threshold was measured. 20 participants then returned for three sessions on separate days, during which the encoded memory was reactivated with only five near-threshold reminder trials (Reactivation group). The additional 20 participants performed full standard daily sessions (Full-Practice group). A standard retest session was performed on Day5 to measure the final discrimination thresholds. Pre- and post-learning fMRI scans were performed on Day1 and Day5. Results show that both groups exhibited significant and comparable learning. A whole-brain analysis showed that a widespread network of higher-order attentional regions is activated while performing the task pre-learning. Following reactivation-induced learning, activity relative to baseline in the intra-parietal sulcus (IPS) and the precuneus was greater compared to full-practice learning. Accordingly, learning-induced changes in inter-regional connectivity are measured using resting-state fMRI. These results suggest that transitioning from repetition-based to reactivation-induced perceptual learning may minimize practice duration and the impact of low-level implicit processing, and engage higher-order attentional resources, potentially beneficial for enhancing generalization of learning.