Abstract
Perceptual learning (PL) is a manifestation of neural plasticity and is known to last for months. Some studies have shown that PL increased activity in the human primary visual cortex (V1) shortly after the behavioral training. However, it is not clear how activity changes through/after the consolidation period. We measured BOLD activity at four different phases in 3 weeks of time course of PL. Six volunteers participated in a series of sessions with visual texture discrimination task (Karni & Sagi, 1991). In training sessions, the target texture was presented only at the upper left visual field and only behavioral performance was measured. In fMRI sessions, the target texture was presented either at the trained visual quadrant (trained condition) or at the untrained visual quadrant (control condition). Four fMRI sessions were scheduled (1) before the first training session (pre-training), (2) on the next day of the first training session (post-1), (3) on the next day of the 7th training session (post-2), and (4) approximately 2 weeks after the post-2 scan (post-3). We compared BOLD activity in V1 between the trained and control conditions. First, no significant difference between these conditions was found at the pre-training scan. Second, consistent with the previous studies, in the trained condition V1 activity was significantly higher at the earlier phase of PL (post-1 and post-2). However, to our surprise, while the behavioral performance sustained after post-2 scan, at post-3 scan, V1 activity decreased to the same level as in the pre-training sessions. The parietal region showed no significant difference in the time course, ruling out the possibility of attentional artifact. Our results suggest that the increased BOLD activity caused by PL is just a transient state of neural plasticity and that downscaling neural activity occurs with reorganization of visual cortex for consolidation of PL.
This study is funded by National Institutes of Health (R01 EY015980, R21 EY017737) to TW the ERATO Shimojo Implicit Brain Function Project to YS. MRI was supported by National Center for Research Resources P41RR14075, the Mental Illness and Neuroscience Discovery Institute, the Athinoula A. Martinos Center for Biomedical Imaging.