Visual mismatch negativity and unintentional temporal-context-based prediction in vision.

Since the discovery of an event-related brain potential (ERP) component, auditory mismatch negativity (auditory MMN), there has been a long-lasting debate regarding the existence of its counterparts in other sensory modalities. Over the past few decades, several studies have confirmed the existence of mismatch negativity in the visual modality (visual MMN) and have revealed the various characteristics of visual MMN. In the present review, a full range of visual MMN studies are overviewed from the perspective of the predictive framework of visual MMN recently proposed by Kimura et al. (2011b). In the first half, the nature of visual MMN is reviewed in terms of (1) typical paradigm and morphologies, (2) underlying processes, (3) neural generators, and (4) functional significance. The main message in this part is that visual MMN is closely associated with the unintentional prediction of forthcoming visual sensory events on the basis of abstract sequential rules embedded in the temporal context of visual stimulation (i.e., "unintentional temporal-context-based prediction in vision"). In the second half, the nature of the unintentional prediction is discussed in terms of (1) behavioral indicators, (2) cognitive properties, and (3) neural substrates and mechanisms. As the main message in this part, I put forward a hypothetical model, which suggests that the unintentional prediction might be implemented by a bi-directional cortical network that includes the visual and prefrontal areas.