Anticipatory Cortical Activation Precedes Auditory Events in Sleeping Infants

Background: Behavioral studies have shown that infants can form associations between environmental events and produce anticipatory actions for the predictable event, but the neural mechanisms for the learning and anticipation of events in infants are not known. Recent neuroimaging studies revealed that the association cortices of infants show activation related to auditory-stimulus discrimination and novelty detection during sleep. In the present study, we expected that when an auditory cue (beeps) predicted an auditory event (a female voice), specific regions of the infant cortex would show anticipatory activation before the event onset even while sleeping. Methodology/Principal Findings: We examined the cortical activation of 3-month-old infants during delays between the cue and the event by using multi-channel near-infrared spectroscopy. To investigate spatiotemporal changes in cortical activation over the experimental session, we divided the session into two phases (early and late phase) and analyzed each phase separately. In the early phase, the frontal regions showed activation in response to the cue that was followed by the event compared with another cue that was not followed by any event. In the late phase, the temporoparietal region, in addition to the frontal region, showed prominent activation in response to the cue followed by the event. In contrast, when the cue was followed by an event and no-event in equal proportions, cortical activation in response to the cue was not observed in any phase. Conclusions: Sleeping 3-month-old infants showed anticipatory cortical activation in the temporoparietal and frontal regions only in response to the cue predicting the event, suggesting that infants can implicitly form associations between temporally separated events and generate the anticipatory activation before the predictable event. Furthermore, the different time evolution of activation in the temporoparietal and frontal regions suggests that these regions may be involved in different aspects of learning and predicting future events.