Figure 1.
Stimulus presentation sequence.
Each trial lasted 2015 ms and consisted of: a central fixation (260 ms) period, followed by the presentation of the stimulus (130 ms) and a second fixation period (260 ms) before a response period (1365 ms) in which the observers (Os) had to give a left or right key press response. Os fixated centrally throughout each trial, although the central fixation point temporally disappeared with the appearance of the stimulus. The stimulus consisted of three Gabor patches (3 cpd, σ = 0.43): two flanking patches forming a 45° reference line and a central target patch. Time flows from right to left. Interpolation task: Os judged whether the central target was offset above and to the left or below and to the right of the 45° reference line. Orientation task: Os judged whether the carrier grating of the central target patch was oriented clockwise or anticlockwise relative to the carrier orientation of the flankers. Sensorimotor control task: Os maintained fixation during stimulus presentation and alternately pressed left and right keys on successive trials.
Figure 2.
Experimental conditions compared to fixation.
Clusters activated (relative to fixation) from the random effects group analysis [P<0.001 (uncorrected); see Table 1 for activated foci] are shown for each condition: (A) interpolation, (B) orientation and (C) sensorimotor control on rendered images of a standard brain [Montreal Neurological Institute (MNI)]. Note the lack of posterior parietal activity in the sensorimotor control condition (C) compared to the extensive activations in the two task conditions (A&B), which is focused along (though not restricted to) the intraparietal sulcus. Ringed in red is a region of the occipito-temporal cortex (possibly the lateral occipital complex) that is present in all three conditions, and though only shown in one hemisphere here, was actually activated bilaterally (see Table 1).
Figure 3.
Experimental conditions compared to fixation.
Clusters activated (relative to fixation) in the random effects group analysis are shown here for all conditions [left column; threshold of P<0.001 (uncorrected); see Table 1 for activated foci] along with plots of % BOLD signal change relative to fixation (centre column; group data) and the corresponding averaged time-series (right column) for the following key activated regions: (A) an occipito-temporal (OT) region, (B) the intraparietal sulcus (IPS), and (C) frontal eye fields (FEFs). Plots represent activity combined for each cluster across the two hemispheres relative to fixation. For the intraparietal sulcus (B) clusters of activity were divided into three groups bilaterally, generating three sub-regions of interest (red arrows) moving from posterior regions of the IPS towards more anterior regions. In the plot the posterior region is shown in bright yellow, the middle region in black and the more anterior region in light yellow / grey. The z coordinate (MNI) is given for each horizontal slice shown. Conditions: [Inter, interpolation; Orient, orientation; Sens, sensorimotor control]. *** P<0.001 (one-way ANOVA with Tukey's post hoc analysis).
Figure 4.
Activated clusters for the two critical contrasts: (A&B) interpolation versus orientation and (C&D) orientation versus interpolation are shown from the random effects group analysis [P<0.001 (uncorrected); see Table 2 for full list of activated foci]. In (A) small activated clusters (1–2 voxels in size) are seen in the left fusiform gyrus and left inferior occipital gyrus. In (B) activations are seen in the right inferior and right middle occipital gyri. However, the middle occipital gyrus cluster (upper cluster) is not due to activity in the interpolation condition but a de-activation in the orientation condition relative to fixation. (See text and Figure 5 for information on task-specific deactivations). In the orientation versus interpolation contrast activations are seen in an anterior region of the left fusiform gyrus (C), and bilaterally in the frontal eyefields [red arrow-heads (D)] and caudal regions of the intraparietal sulcus [red arrows (D)]. MNI coordinates are given for each slice presented.
Figure 5.
Deactivations relative to fixation.
Clusters of relative activation / deactivation from the random effects group analysis [see Table 2 for foci; P<0.001 (uncorrected) threshold] are shown along with plots of % BOLD signal change relative to fixation (centre column; group data) and the corresponding averaged time-series (right column) from the following contrasts: (A) interpolation versus orientation and (B) fixation versus all conditions. Several clusters of relative activity in the interpolation versus orientation contrast were found to be due to de-activations in the orientation condition (% BOLD signal change relative to fixation). Significant clusters include the primary auditory cortex bilaterally (A1: red ring) and the middle temporal gyrus bilaterally (GTm: red arrows). For the GTm, clusters of activity were divided into three groups bilaterally, generating three sub-regions moving from posterior, through central and on towards more anterior regions (yellow, black and light yellow / grey plots respectively). The only region showing deactivations (relative to fixation) in all conditions was a large midline cluster pertaining to the pre-cuneus / cuneus (Cu: red arrow-head), a well-documented component of the ‘default network’. (See text for details). Conditions: [Interp, interpolation; Orient, orientation; Sens, sensorimotor control]. ** P<0.01; *** P<0.001 (one-way ANOVA with Tukey's post hoc analysis).
Table 1.
Experimental conditions versus fixation.
Table 2.
Direct comparisons of experimental conditions.