Simultaneous selection by object-based attention in visual and frontal cortex

Models of visual attention hold that top-down signals from frontal cortex influence information processing in visual cortex. It is unknown whether situations exist in which visual cortex actively participates in attentional selection. To investigate this question, we simultaneously recorded neuronal activity in the frontal eye fields (FEF) and primary visual cortex (V1) during a curve-tracing task in which attention shifts are object-based. We found that accurate performance was associated with similar latencies of attentional selection in both areas and that the latency in both areas increased if the task was made more difficult. The amplitude of the attentional signals in V1 saturated early during a trial, whereas these selection signals kept increasing for a longer time in FEF, until the moment of an eye movement, as if FEF integrated attentional signals present in early visual cortex. In erroneous trials, we observed an interareal latency difference because FEF selected the wrong curve before V1 and imposed its erroneous decision onto visual cortex. The neuronal activity in visual and frontal cortices was correlated across trials, and this trial-to-trial coupling was strongest for the attended curve. These results imply that selective attention relies on reciprocal interactions within a large network of areas that includes V1 and FEF.Visual scenes are usually too complex for all information to be analyzed at once. Selective attention selects a subset of the objects in the visual scene for detailed analysis at the expense of other items. Visual objects compete for selection, and the outcome of this competition depends on bottom-up cues such as saliency and perceptual organization and top-down cues that signal the objects’ behavioral relevance (1). It is not well understood how these different cues interact and which brain areas take the lead in visual selection.The top-down mechanisms for attentional selection are tightly linked to those for the selection of actions (2), and accordingly, cortical areas related to action planning influence the deployment of visual attention. The frontal eye fields (FEF) is one such area that is involved in visual processing, shifts of visual attention (2, 3), and also in the control of eye movements (4, 5). Area FEF contains different types of cells. Visual processing relies on visual and visuomovement cells, whereas the programming of eye movements relies on the activity of visuomovement and movement cells (6, 7). There are several lines of evidence that also implicate FEF in attentional control. First, FEF inactivation impairs attention shifts toward the contralateral visual field (8, 9). Second, subthreshold FEF microstimulation enhances neuronal activity in visual cortex in a manner that is reminiscent of selective attention (10, 11). Third, a role of FEF in the top-down guidance of attention is supported by studies on visual search. In search, selection signals in frontal cortex precede those in area V4 by 50 ms, suggesting that the frontal cortex determines selection and then provides feedback to visual cortex (12, 13). A comparable interareal delay in attentional effects was observed between the lateral intraparietal area and the motion sensitive middle temporal area (14). Thus, the parietal and frontal cortices appear to take the lead in attentional selection and to provide top-down signals to visual cortex. Within the visual cortex, such a reverse hierarchy (15) of attentional effects was observed in a task that required shifts of spatial attention (16) and also in a task demanding shifts between visual and auditory attention (17). Attentional signals in area V4 preceded signals in V2 by 50–250 ms, which in turn preceded attentional effects in the primary visual cortex (V1) by 50–400 ms.However, top-down factors are not the only ones that guide attention. Attention can be object-based, implying that the visual stimulus itself influences the distribution of attention too. If attention is directed to a feature, attention tends to coselect visually related features on the basis of perceptual grouping cues (18) so that entire objects rather than isolated features are attended (19, 20). The influence of perceptual grouping on attentional selection can be investigated with a curve-tracing task that requires grouping of the contour elements of a single curve (21, 22). Attention in this task is directed to the entire curve, implying that the curve’s shape itself influences the distribution of attention (22). Indeed, a traced curve evokes stronger activity in primary visual cortex than an irrelevant curve, revealing a neuronal correlate of object-based attention (23). However, it is not known if the coselection of all image elements of a single object is determined within early visual cortex or is guided by the frontal cortex, just as was shown for other tasks.Here we compare selection signals in areas FEF and V1 in the curve-tracing task with simultaneous recordings in the two areas. A priori, several possibilities exist for the interaction between V1 and FEF. First, the frontal cortex might select the relevant curve and then feed a guiding signal back to visual cortex (24, 25) as in the other tasks described above. If so, attentional selection signals in V1 might arise tens to hundreds of milliseconds later than in FEF. However, the chain of events in the curve-tracing task might differ because visual shape has a profound influence on the distribution of attention (26). Thus, a second possibility is that the visual cortex determines selection so that the attentional modulation in visual cortex precedes that in frontal cortex. A third possibility is that visual and frontal areas jointly determine what is relevant and what is not. In this situation, the selection signals are expected to occur in both areas at approximately the same time. It is also possible that the order of selection in different areas depends on the difficulty of the task. For example, the reverse hierarchy theory of visual perception (15) proposed that easy tasks are usually solved by higher visual areas, whereas lower visual areas are recruited when the picture has to be scrutinized. We therefore varied the difficulty of the curve-tracing task.