It then takes a few tens of milliseconds before FGM also emerges in the center of the figure in V1 (Lamme et al., 1999). The effects of attention are observed

at buy Afatinib yet later time points; attention first increases FGM in V4 and it then also boosts center modulation in V1 (Ogawa and Komatsu, 2006 and Roelfsema et al., 2007). One must be cautious when inferring connectivity from latency differences alone. For example, the effect of feedback to V1 may under some conditions be faster than influences caused by horizontal connections (Bair et al., 2003). However, the difference between the mechanisms for edge- and center-FGM is supported by a number of additional observations. First, task-driven attention boosted the representation

of the figure center and had less effect on the edge representation (Figure 8E). This implies that edge-FGM is largely stimulus driven, whereas center-FGM depends more on feedback from higher areas. Second, a previous study (Lamme et al., 1998a) showed that lesions in higher Ixazomib research buy visual areas reduce center FGM in V1 but leave edge modulation intact (see also Hupé et al., 1998). Third, we could reproduce the timing and the spatial profile of the visual responses, the FGM and the attentional modulation in V1 and V4 with a model that detects boundaries with local inhibition and uses excitatory feedback for region filling. These results imply that the mechanisms proposed by us are sufficient to explain the data. The enhancement of neuronal activity at boundaries occurs quickly (Lamme et al., 1999 and Nothdurft et al., the 2000) and is not strongly modulated by attention. Previous studies demonstrated that texture elements surrounded by dissimilar elements are more salient

(Joseph and Optican, 1996). Image elements that pop out cause stronger neuronal activity in visual cortex during an early response phase (Burrows and Moore, 2009, Kastner et al., 1997, Knierim and van Essen, 1992, Lamme et al., 1999, Lee et al., 2002, Nothdurft et al., 1999 and Ogawa and Komatsu, 2006) and a similar increase in V1 activity occurs at the location of an edge where the orientation changes abruptly (Nothdurft et al., 2000). These saliency effects also occur when animals ignore the stimulus (Knierim and van Essen, 1992) (but see Burrows and Moore, 2009), and even if they are anesthetized (Kastner et al., 1997, Nothdurft et al., 1999 and Nothdurft et al., 2000). Accordingly, image elements can pop out in psychophysics (Theeuwes et al., 2006) if they are not relevant to the task, although these effects are transient and disappear after 250 ms (Donk and van Zoest, 2008 and Joseph and Optican, 1996). It is likely that edge-FGM is related to neuronal responses in V1, V2, and V4 that reflect the assignment of the edge to the figural side, because borders “belong” to figures and not to the background (Zhou et al., 2000).