Direct evidence for attention-dependent influences of the frontal eye-fields on feature-responsive visual cortex

Abstract

Voluntary selective attention can prioritize different features in a visual scene. The frontal eye-fields (FEF) are one potential source of such feature-specific top-down signals, but causal evidence for influences on visual cortex (as was shown for "spatial" attention) has remained elusive. Here, we show that transcranial magnetic stimulation (TMS) applied to right FEF increased the blood oxygen level-dependent (BOLD) signals in visual areas processing "target feature" but not in "distracter feature"-processing regions. TMS-induced BOLD signals increase in motion-responsive visual cortex (MT+) when motion was attended in a display with moving dots superimposed on face stimuli, but in face-responsive fusiform area (FFA) when faces were attended to. These TMS effects on BOLD signal in both regions were negatively related to performance (on the motion task), supporting the behavioral relevance of this pathway. Our findings provide new causal evidence for the human FEF in the control of nonspatial "feature"-based attention, mediated by dynamic influences on feature-specific visual cortex that vary with the currently attended property.

Keywords: concurrent TMS-fMRI; feature attention; frontal eye-fields; top-down control.

Figure 1.

Schematic example of a single trial. Participants were instructed either to attend the gender of a face (as indicated by “f” cue) or the motion direction (“m” cue) in a field of dots (containing 20% random motion) or passively view the display (“p” cue), with both types of stimuli superimposed on displays that were visually equivalent across all conditions. Short bursts of transcranial magnetic stimulation (TMS) (3 pulses at 11 Hz at either 40% or 110% of RMT) were applied over right FEF, commencing 40 ms following target display onset. ITI, intertrial interval (2.5–3.5 volume repetitions [TR] or 8910 ± 1485 ms). CTI, cue-target interval. The TMS parameters were selected based on other recent TMS-fMRI studies (Feredoes et al. 2011; Heinen et al. 2011) and TMS safety guidelines (Rossi et al. 2009).

Figure 2.

TMS effects on BOLD signal in the stimulated right FEF and the left FEF ROI, depend on active attention. (a) The brain image depicts the mean location, averaged across participants, of 10 mm spheres (shown in red) used as ROIs for the targeted right FEF, and for the left FEF, projected onto a MNI normalized brain (mean MNI coordinates right FEF: 31, 1, 58; mean MNI coordinates left FEF: −31 −3 57). (b) The bars display the TMS-induced difference (high minus low intensity) in BOLD signal extracted from the right FEF ROI for all 3 tasks conditions. TMS affected the BOLD signal for the motion and face tasks, but not for the passive viewing task (see main text). Error bars show ±SEM, and asterisks indicate significant (P < 0.05) differences in post hoc paired t-tests. See Supplementary Figure 2a for BOLD signals for all 6 conditions separately. (c) A similar pattern to right FEF-TMS effects on BOLD activity was found in the left FEF ROI (see also Supplementary Fig. 2b).

Figure 3.

Remote effects of right FEF TMS on BOLD signal for MT+ and FFA are specific to the attended feature. (a) The left panel shows the group ROI for MT+ (centered on MNI coordinates −45, −70, 4 for left hemisphere and 46, −61, 2 for right) defined jointly by the Motion > Face task contrast and anatomical constraints (see Materials and Methods section). The plot to the right reveals significant TMS effects (high minus low TMS intensity) in MT+ (pooled across hemispheres) for attend-motion but not for attend-face or passive viewing conditions. Error bars are ±SEM and asterisks indicate significant (P < 0.05) differences in pairwise t-tests. See Supplementary Figure 2c for BOLD signal for all 6 conditions separately. (b) The left panel displays a compound image combining all individual FFA ROIs per participant (centered on MNI coordinates of −35 −54 −14 for left hemisphere, 36 −43 −17 for right) as defined by the Face > Motion contrast for each participant individually together with anatomical constraints (see Materials and Methods section). TMS effects (high minus low intensity) were observed in FFA (averaged across right and left hemispheres) for the attend-face condition, but not for attend-motion or passive viewing conditions. See Supplementary Figure 2d for BOLD signal for all 6 conditions separately.

Figure 4.

TMS effect on performance and its relation to TMS effects on BOLD in right FEF and MT+ for the motion task. (a) Mean IE scores (see Materials and Methods section) are shown separately for the high- and low-intensity FEF TMS in the motion and face task, with error bars showing ±1 SEM. High TMS led to worse performance (higher IE) on the motion task. Performance on the face task was overall better than on the motion task, and there was no significant effect of effective versus ineffective TMS on the face task. (b) Scatterplot, with regression line, for TMS-induced differences in BOLD signal in MT+, plotted against TMS-induced differences in IE scores for the motion task. Each data point represents one participant. Larger TMS-induced BOLD increases in MT+ were correlated with larger decrements in performance for the motion task. (c) A similar relationship was found in right FEF, that is, larger TMS-induced increases in right FEF BOLD signal correlated with larger performance decrements during the motion task.