Additionally, by alternating blocks in which the animals needed to detect orientation and spatial frequency changes they could compare responses selleck compound when one or the other feature was attended and isolate the effects of feature-based attention ( McAdams and Maunsell, 2000). The authors found that populations of V4 neurons could independently show both types of attentional modulation. For example, a neuron could be modulated by spatial attention but not by feature-based attention and vice-versa. One main difference between the effects of spatial and feature-based attention was that the former enhanced responses of neurons within the hemisphere

contralateral to the attended stimulus, while the Akt inhibition latter enhanced neuronal responses in both hemispheres, irrespective of the attended stimulus location. The feature-based attentional modulation was dependent on the relationship between the attended stimulus feature and the cell’s preferred feature (FSG, see Figure 2 of Cohen and Maunsell [2011]). For example, the response of a neuron when animals attended to a particular orientation was enhanced if the unit preferred that orientation but was suppressed if the attended orientation was antipreferred. FSG, as opposed to FM, produces enhanced or suppressed responses in neurons

with receptive fields containing stimuli with the target features, depending the on the units’ feature selectivity

(Treue and Martínez Trujillo, 1999). Moreover, recording from 96 electrodes at a time (48 in each hemisphere) allowed the authors to examine the impact of spatial and feature-based attention on spike count correlations, a variable that has been shown to be influenced by the allocation of attention (Cohen and Maunsell, 2009 and Mitchell et al., 2009). V4 units showing increases in response by both spatial- and feature-based attention show decreases in correlation, while V4 units showing response decreases by either type of attention showed increases in correlation. This suggests that response modulation and correlated firing are two sides of the same coin. Any variable that increases or decreases the firing rate of visual neurons to sensory stimuli (e.g., changes in contrast or adaptation) will likely produce decreases or increases in correlated firing, respectively, and therefore will influence the ability of neuronal populations to encode visual information. Supporting this hypothesis, spike count correlations between pairs of MT neurons decrease when increasing stimulus contrast (Huang and Lisberger, 2009). The exact mechanisms of these effects need to be elucidated.