Examining the Effect of Ketamine on Gestalt Perception

Predictive coding postulates that the organisation of the visual scene not only relies on the incoming information, but also on prior knowledge about the world. As such, perception reflects the combination of bottom-up sensory information with top-down prior beliefs (predictions) about the source of the sensory input. These top-down prior beliefs are thought be fed back from higher hierarchical levels to sensory cortices via glutamatergic N-methyl-D-aspartate receptor (NMDA-R) signalling to be compared with the sensory input. In accord with this, there is evidence that the NMDA-R antagonist ketamine weakens the role of priors in perception. Moreover, there are mixed reports that schizophrenic patients, who have a hypo-function of cortical NMDA-Rs, show greater resistance to visual illusions. Here, we investigated in a placebo-controlled blinded study the effect of ketamine on the perception of an asymmetric bi-stable motion illusion. The stimulus consisted of four pairs of dots that could either be perceived as unbound dots moving locally (low-level default perception) or as two squares sliding over each other in transparent motion (prior-belief dependent high-level perception). Following predictive coding, we hypothesise to observe a relative decrease of the prior-belief dependent perception after ketamine infusion. However, preliminary results reveal no treatment-dependent effects