Brain regions modulated during covert visual attention in the macaque

Abstract

Neurophysiological studies of covert visual attention in monkeys have emphasized the modulation of sensory neural responses in the visual cortex. At the same time, electrophysiological correlates of attention have been reported in other cortical and subcortical structures, and recent fMRI studies have identified regions across the brain modulated by attention. Here we used fMRI in two monkeys performing covert attention tasks to reproduce and extend these findings in order to help establish a more complete list of brain structures involved in the control of attention. As expected from previous studies, we found attention-related modulation in frontal, parietal and visual cortical areas as well as the superior colliculus and pulvinar. We also found significant attention-related modulation in cortical regions not traditionally linked to attention - mid-STS areas (anterior FST and parts of IPa, PGa, TPO), as well as the caudate nucleus. A control experiment using a second-order orientation stimulus showed that the observed modulation in a subset of these mid-STS areas did not depend on visual motion. These results identify the mid-STS areas (anterior FST and parts of IPa, PGa, TPO) and caudate nucleus as potentially important brain regions in the control of covert visual attention in monkeys.

Figure 1

Behavioral tasks and performance. The color of the square around the central fixation spot instructed the monkey to either monitor the fixation stimulus (black in baseline (a) and Ignore (b) tasks) or the peripheral motion stimuli (red in Attend (c) task). (a) Baseline task. Following 500 ms of fixation, the central fixation spot dimmed on 65% of the trials during the delay period. Monkey released the joystick within 0.8 seconds of the dim to get a juice reward. (b) Performance of both monkeys (Monkey #1: blue; Monkey #2: green) on change trials (dim) and catch trials (no dim) in the Baseline task. Circles indicate % response in each session. Horizontal black lines with gray bars indicate mean and standard deviation respectively. (c) Ignore task. Following 500 ms of fixation, two circular patches of random dot motion stimuli (6° in diameter) appeared on either side of fixation at 8° eccentricity (radius) and 10° above horizontal (azimuth). During the delay period, the central fixation spot dimmed on 65% of the trials and monkey released the joystick within 0.8 seconds of the dim to get a juice reward. During the same delay period, independently of the fixation spot dimming, one of the motion stimuli changed direction on 65% of the trials. Monkey had to ignore the motion-change and hold the joystick down. (d) Performance in the Ignore task. Color and symbol conventions same as (b). (e) Attend task. Following 500 ms of fixation, two circular patches of random dot motion stimuli appeared at the same location as in Ignore task. During the delay period, one of the motion stimuli changed direction on 65% of the trials and monkey released the joystick within 0.8 seconds of the motion-change event to get a juice reward. (f) Performance in the Attend task to left and right motion-changes as well as no changes. Color and symbol conventions same as (b). (g) Block Design: All three tasks were presented in a block design and the duration of each task in the block design is shown in d. Each run started with the Baseline task and was interleaved with Ignore and Attend tasks.

Figure 2

Activations during Ignore and Attend tasks. (a,b) T-scores contrasting Ignore and baseline tasks show activations during Ignore task in left (a) and right (b) hemispheres of monkey #1. (c,d) T-scores contrasting Attend and baseline tasks show activations during Attend task in left (c) and right (d) hemispheres of monkey #1. Anatomical boundaries are labeled for the left hemisphere. T-scores were corrected for multiple comparisons (Bonferroni correction; p < 0.05, |t-score| > 5.02).

Figure 3

Cortical maps of attention-related activation. T-scores contrasting Attend and Ignore tasks were projected onto inflated cortical surfaces of D99 in each monkey’s native space along with anatomical boundaries (black contours). (a,b) Inflated cortical maps of t-scores showing attention-related activation in left (a) and right (b) hemispheres of monkey # 1. Anatomical boundaries are labeled for the left hemisphere in monkey # 1. (c,d) Inflated cortical maps of t-scores showing attention-related activation in left (c) and right (d) hemispheres of monkey # 2. Anatomical boundaries are labeled for the left hemisphere in monkey # 2. T-scores were corrected for multiple comparisons (Bonferroni correction; p < 0.05, t-score > 5.02).

Figure 4

Attention-related modulation in mid-STS areas. (a,b) Magnified versions of Fig. 3a,b showing STS activation in left and right hemispheres of monkey # 1. (c,d) Magnified versions of Fig. 3c,d showing STS activation in left and right hemispheres of monkey # 2. The white arrow points to the peak activation in the aFST/IPa region in left and right hemispheres of both monkeys.

Figure 5

BOLD time-courses during Attend and Ignore tasks in cortical ROIs. (a–i) Each row shows plots of mean BOLD time-courses for a given area in left and right hemispheres of both monkeys (Monkey # 1: Left column; Monkey # 2: Right column) along with coronal slices containing the peak of the area. The name of the area is shown at top-left of each row. The location of the coronal slice w.r.t the inter-aural axis is shown on top of each coronal slice. Mean BOLD time-courses are plotted as % change in BOLD on y-axis against repetition time (TR) on the x-axis for Attend (red) and Ignore (blue) tasks. Error bars indicate 95% confidence intervals.TR = 0 on x-axis indicates start of the block. The location of the white cross-hair in each coronal slice indicates the peak of the area in the corresponding hemisphere and is overlaid with attention-related activation (t-scores contrasting Attend and Ignore (Bonferroni correction; p < 0.05, t-score > 5.02), as in Fig. 3). aFST refers to anterior part of anatomical area FST.

Figure 6

BOLD time-courses during Attend and Ignore tasks in subcortical ROIs. (a–c) Same conventions as Fig. 5.

Figure 7

Second-order orientation stimulus: behavioral tasks and performance. (a) Second-order orientation stimulus. A 2D uniform distribution centered at zero was added to a base luminance (c) to generate a white noise stimulus. The same 2D uniform distribution was modulated by a 2D sinusoid and added to a base luminance to generate a second-order orientation stimulus. The orientation seen in the stimulus is second-order, because it is based on the local contrast of the bands of the grating without any difference in the local mean luminance of the bands of the grating. (b–d) All task conventions were the same as in Fig. 1, except that the peripheral stimuli were second-order orientation patches rather than visual motion patches. (e) Block Design: All three tasks were presented in a block design identical to that used in the motion version of the task (Fig. 1).

Figure 8

Cortical areas showing attention-related modulation to second-order orientation stimulus. (a,b) T-scores contrasting Attend and Ignore tasks described in Fig. 7 were projected onto inflated cortical surfaces of D99 in native space of monkey # 1 along with anatomical boundaries (black contours). Inflated cortical maps of t-scores (Bonferroni correction; p < 0.05, t-score > 5.02) showing attention-related activation in left (a) and right (b) hemispheres of monkey # 1. Anatomical boundaries for the left hemisphere of monkey # 1 are shown in Fig. 3a. (c) % Overlap for a given area was defined as the percentage of total voxels in that area that showed significant attention-related modulation in both motion (Fig. 1) and second-order orientation (Fig. 7) versions of the tasks.

Figure 9

Overlap of attention-related modulation in mid-STS areas during orientation and motion tasks. (a,b) Magnified versions of Fig. 8a,b showing STS activation during orientation task in left and right hemispheres. Blue contour represents the contiguous activation in mid-STS areas. (c,d) Magnified versions of Fig. 3a,b showing STS activation during motion task in left and right hemispheres. Blue contours from (a,b) were overlaid for the left and right hemispheres respectively.