Research ReportPre-target activity in visual cortex predicts behavioral performance on spatial and feature attention tasks
Introduction
One of the key functions of the attention system is to aid the selection and processing of task-relevant information in our environment in support of coherent goal-directed behavior. A long line of behavioral, neurophysiological, and neuroimaging studies of visual selective attention have demonstrated that voluntary covert orienting (i.e., orienting without eye movements) to a particular spatial location or to a particular nonspatial stimulus feature (e.g., color) leads to facilitated behavioral performance and to increased neuronal responses evoked by stimuli that are presented in the attended location or that possess the attended feature (e.g., Corbetta et al., 1990, Heinze et al., 1994, Hillyard and Münte, 1984, Kingstone and Klein, 1991, Luck et al., 1994, Mangun and Hillyard, 1991, Posner, 1980, Van Voorhis and Hillyard, 1977, Woldorff et al., 1997). Neurobiological models of attention posit that the enhanced behavior and associated neuronal activity reflect the influence of top–down control mechanisms on bottom–up sensory processing in visual cortex (e.g., Corbetta and Shulman, 2002, Desimone and Duncan, 1995).
Although attention effects in sensory processing are typically observed as modulations of the neural activity evoked by the task-relevant “target” stimuli, several studies have observed modulations in neural activity prior to, or in the absence of, the to-be-attended target stimulus. For example, single-unit recordings in monkeys have found that cells in inferior temporal cortex that are selective for a given stimulus exhibit elevated pre-stimulus baseline firing rates when that stimulus is to be fixated in an upcoming visual search display (Chelazzi et al., 1993, Chelazzi et al., 1998). Similarly, Luck et al. (1997) found that some cells in visual cortex exhibit increases in pre-stimulus baseline firing rates when the visual field location that the cells represent is covertly attended. As in the single-unit studies, human neuroimaging studies have also reported increases in activity prior to the presentation of an attended target stimulus in areas of visual cortex that represent the attended location (e.g., Giesbrecht et al., 2003, Hopfinger et al., 2000, Kastner et al., 1999, Weissman et al., 2004, Wilson et al., 2005, Woldorff et al., 2004) or the attended feature (Chawla et al., 1999).
According to one model of visual selective attention, changes in pre-stimulus activity are generated by top–down biasing signals originating from control areas in frontal and parietal cortex (Desimone and Duncan, 1995). As a consequence of these biasing signals, the attended stimulus is afforded an advantage in the competition for neural resources, resulting in a higher level of processing of the attended items relative to unattended ones. If changes in baseline activity do indeed bias processing in favor of the attended stimulus, then it follows that these changes in baseline activity should be related to behavioral performance. Consistent with this conjecture, studies of spatial attention have reported that activity in visual cortex is correlated with the outcome of perceptual decisions in highly trained observers performing a near-threshold detection task (Ress and Heeger, 2003, Ress et al., 2000). Although consistent with the theoretical conjecture, it remains unclear whether these correlations between activity in visual cortex and behavior generalize to cue-triggered, pre-stimulus modulations of neural activity and to the behavior of naive participants during suprathreshold discrimination tasks that are more commonly used when studying visual selective attention.
The present study had two main goals. First, we investigated the magnitude and time course of pre-stimulus activity in regions of visual cortex, in response to cues that directed subjects to attend to a spatial location or to a stimulus feature prior to target presentation. Although there is evidence that attending to spatial locations or nonspatial stimulus features can result in enhanced pre-stimulus activity, the evidence comes from different studies using a variety of tasks. Therefore, it has been difficult to relate pre-stimulus modulations during spatial and feature modes of attention (but see Giesbrecht et al., 2003). Second, and most importantly, we investigated whether pre-stimulus modulations that occurred during cued spatial and cued feature attention predicted subsequent behavioral discrimination performance of a target that was presented in the cued location or possessed the cued feature.
To address these questions, we asked participants to perform a cued attentional orienting task, while their brain activity was assessed using functional magnetic resonance imaging (fMRI). In each trial, participants were cued to covertly direct their attention to either one of two spatial locations (right or left) or one of two colors (blue or yellow). The task was to discriminate the orientation (horizontal or vertical) of a rectangle that was presented at the cued location or in the cued color (Fig. 1). Critically, the spatial and nonspatial cuing conditions were equated on a variety of nonspecific effects (e.g., arousal, motor preparation) and cognitive operations (e.g., working memory, response selection) but differed in terms of which stimulus dimension was cued. We predicted that location- and color-selective regions of the visual cortex would both exhibit selective increases in pre-target activity in response to cues to attend for the appropriate target. Moreover, we hypothesized that these increases in pre-target activity would be associated with improved performance on the location- and color-cue tasks. Consistent with these predictions, we observed activation increases in visual cortical areas selective for processing the target stimuli prior to their actual presentation, and the magnitude of these modulations was positively correlated with behavioral performance across subjects.
Section snippets
Behavior
The mean proportion of correct responses in the spatial and color conditions is shown in Fig. 2. Overall, subjects performed the tasks well, with mean proportion correct being 0.85. Subjects were more accurate on the color task (t(11) = 9.35, P < 0.001), where the mean proportion correct was 0.867 (SEM = 0.003), than on the location task, where the mean proportion correct was 0.837 (SEM = 0.003).
Pre-target modulations
The areas that exhibited differential increases in pre-target activity that were significantly
Discussion
The purpose of the present investigation was to investigate (a) pre-target modulations of activity during cued selective attention and (b) the relationship between pre-target modulations during cued selective attention and subsequent behavioral performance on the attention task. We hypothesized that areas of visual cortex that selectively responded to the location and feature targets would exhibit increases in activity in response to cues to attend to a location or a feature, but before the
Conclusion
A central issue in recent neuroimaging studies of visual attention has been to identify the brain systems and mechanisms that control selective attention. Emerging from these studies is the view that portions of frontal and parietal cortex are involved in the control of attentional orienting to spatial locations, spatial reference frames, objects, and other nonspatial stimulus features (Corbetta et al., 2000, Giesbrecht et al., 2003, Giesbrecht et al., 2005, Hopfinger et al., 2000, Shulman et
Subjects
Sixteen subjects (8 female; ages 24–32) gave informed consent before participating in accordance with the guidelines of the local Institutional Review Board. The data from ten of these subjects were the subject of a previously published article addressing a different theoretical question (Giesbrecht et al., 2003).
Stimuli
Cues were gray uppercase letters from the English alphabet presented at fixation and were 0.8 × 0.6° (height × width). The targets were rectangles (1.75° × 1.42°) that could be
Acknowledgments
This research was supported by NIMH grant R01MH55714 (GRM, BG), NINDS grant P01-NS41328-Proj. 2 (MW, GRM), the McDonnell-Pew Program in Cognitive Neuroscience (BG), and the Natural Sciences and Engineering Research Council of Canada (BG). We thank Jocelyn Sy and David Horton for their helpful comments on earlier versions of the manuscript.
References (45)
- et al.
Functional magnetic resonance imaging (FMRI) of the human brain
J. Neurosci. Methods
(1994) - et al.
Event-related fMRI: characterizing differential responses
NeuroImage
(1998) - et al.
Neural mechanisms of top–down control during spatial and feature attention
NeuroImage
(2003) - et al.
Increased activity in human visual cortex during directed attention in the absence of visual stimulation
Neuron
(1999) - et al.
Combining spatial extent and peak intensity to test for activations in functional imaging
NeuroImage
(1997) - et al.
A role for top–down attentional orienting during interference between global and local aspects of hierarchical stimuli
NeuroImage
(2002) - et al.
Control networks and hemispheric asymmetries in parietal cortex during attentional orienting in different spatial reference frames
NeuroImage
(2005) - et al.
Randomized event-related experimental designs allow for extremely rapid presentation rates using functional MRI
NeuroReport
(1998) - et al.
The physiological basis of attentional modulation in extrastriate visual areas
Nat. Neurosci.
(1999) - et al.
A neural basis for visual search in inferior temporal cortex
Nature
(1993)