The effect of short-term training on repetition probability effects for non-face objects.

The magnitude of repetition suppression (RS) is modulated by the probability of stimulus repetitions when measured by fMRI. This repetition probability effect (P(rep)) is usually expressed in a stronger RS when the probability of repetition trials is higher when compared to blocks with less frequent repetitions. Previous studies have suggested that the P(rep) effect depends on the extensive long-term experience of participants with the stimuli. At the same time, the how short-term learning affects RS and its modulation by P(rep) remains largely unknown. To address this issue, we used fMRI and measured the RS and its modulation by P(rep) for non-face objects (cars) before and after a 10-day long perceptual learning (PL) period and for unfamiliar faces as control. The results showed a significant P(rep) effect for faces within the Fusiform Face Area (FFA) and for cars within the Lateral Occipital Complex (LO) in the pre-training fMRI measurement session. Following the PL period, participants exhibited strong improvements in the subordinate categorization of the trained stimuli. Surprisingly, the magnitude of RS did not change as a function of training, but the P(rep) effect was absent in the post-training fMRI sessions for both stimulus categories. These results suggest that the predictive processes, measured by P(rep) modulation of RS, may be modulated by the short-term perceptual learning experience.