Familiarity breeds order

Abstract

Attention and working memory mechanisms need to have on-line access to acquired information in order to coordinate behaviour. The characterization of neural mechanisms that underlie adult learning by experience can lead to the formulation of models of how the brain uses such information to guide thought and action at the neuronal level. There has been an accumulation of evidence from primate studies that neurons in ‘higher’ visual cortical areas – in particular, the inferior temporal (IT) cortex – adjust their pattern of response according to the level and nature of exposure of the animal to a visual stimulus. In other words, they exhibit a kind of ‘neuronal learning’. Erickson and colleagues 1xClustering of periphinal neurons with similar properties following visual experience in adult monkeys. Erickson, C.A. et al. Nat. Neurosci. 2000; 3: 1143–1148Crossref | PubMed | Scopus (89)See all References recently reported that neurons in the perirhinal cortex (an anteromedial area in IT) show enhanced similarity in their patterns of activity in response to visual input, according to their proximity to each other and the level of familiarity with the encountered object.They presented images of novel objects to macaque monkeys, which were required either to perform a visual discrimination task or to view the stimuli passively. The activity of groups of neurons in the perirhinal cortex was recorded during these presentations. Erickson et al. analysed their results according to the degree of familiarity of the object presented and the proximity of the neurons sampled. They described how ‘nearby’, but not ‘far-apart’ neurons exhibited similar patterns of activity when the animal was presented with an object encountered in a previous session, but when the object was novel, there was no organized similarity between the patterns of activity of neurons, irrespective of their distance. The significance of these results lies in the fact that they concern a high order visual area, believed to be involved in object recognition memory.They provide an example of functional neuronal architecture altered by experience in adulthood, quite distinct from the hard-wired organization in lower (primary) visual areas early in development, where cells are clustered in predetermined columns of preference for different object properties such as line orientation.Although the cellular mechanisms of the experience-dependent alterations described are not known, the results of Erickson et al. add weight to a growing body of evidence that suggests a way in which higher visual areas might contribute to an extended network of perceptual organization. This kind of neuronal learning can be viewed as organizing neuronal nodes of a network of ‘visual templates’ that resolve or bias competition – in this case according to familiarity – between incoming information from the primary visual cortex, for access to high-order centres for behavioural organization. This appears to be a highly viable proposal, both in light of the neurophysio-logical evidence and the nature of the connectivity of the perirhinal cortex with both primary visual areas and frontal high-order systems, such as the prefrontal and cingulate cortices 2xPerceptual–mnemonic functions of the perirhinal cortex. Murray, E.A. and Bussey, T.J. Trends Cognit. Sci. 1999; 3: 142–151Abstract | Full Text | Full Text PDF | PubMed | Scopus (332)See all References.