Posteromedial parietal cortical activity and inputs predict tactile spatial acuity

Abstract

We used functional magnetic resonance imaging (fMRI) to investigate the neural circuitry underlying tactile spatial acuity at the human finger pad. Stimuli were linear, three-dot arrays, applied to the immobilized right index finger pad using a computer-controlled, MRI-compatible, pneumatic stimulator. Activity specific for spatial processing was isolated by contrasting discrimination of left-right offsets of the central dot in the array with discrimination of the duration of stimulation by an array without a spatial offset. This contrast revealed activity in a distributed frontoparietal cortical network, within which the levels of activity in right posteromedial parietal cortical foci [right posterior intraparietal sulcus (pIPS) and right precuneus] significantly predicted individual acuity thresholds. Connectivity patterns were assessed using both bivariate analysis of Granger causality with the right pIPS as a reference region and multivariate analysis of Granger causality for a selected set of regions. The strength of inputs into the right pIPS was significantly greater in subjects with better acuity than those with poorer acuity. In the better group, the paths predicting acuity converged from the left postcentral sulcus and right frontal eye field onto the right pIPS and were selective for the spatial task, and their weights predicted the level of right pIPS activity. We propose that the optimal strategy for fine tactile spatial discrimination involves interaction in the pIPS of a top-down control signal, possibly attentional, with somatosensory cortical inputs, reflecting either visualization of the spatial configurations of tactile stimuli or engagement of modality-independent circuits specialized for fine spatial processing.

Figure 1.

A, MRI-compatible pneumatic stimulator. Stimuli were mounted face-down on the square stage at the bottom of the drive shaft. The finger mold used to immobilize the finger was mounted on the base of the device. Arrows indicate direction of airflow. Disk on the top of the stimulator allowed 180° rotation of the stimulus. B, Stimulus configurations in spatial task; central dot in array was offset either to the right or left. C, Stimulus array for the temporal task used an array without spatial offset.

Figure 2.

Activations in the 22-subject group on the spatial task relative to the temporal task, displayed on representative slices through the anatomic images from one subject using a pseudocolor t scale. Values below each slice indicate its Talairach plane. Activations are derived from a random-effects analysis and FDR corrected for multiple comparisons (q < 0.05). R, Right; A, anterior; IFS, inferior frontal sulcus; MD, mediodorsal; VPL, ventral posterolateral.

Figure 3.

Time courses of BOLD signal change in representative regions that were more active on the spatial task than the temporal task. Abbreviations as in Figure 2. The horizontal (time) axis is calibrated in terms of scan number (TR, 1.5 s). R, Right; L, left.

Figure 4.

Activations on the temporal task relative to the spatial task and their time courses. Details are as in Figures 2 and 3. MFG, Middle frontal gyrus; R, right; L, left.

Figure 5.

Bivariate Granger causality maps using the right pIPS ROI (red) as a reference. Representative slices are illustrated. Top, Better performers. Bottom, Poorer performers. Olive-green ROIs within white ellipses indicate location of medial parietal and right calcarine ROIs selected from these maps for subsequent multivariate Granger causality analyses. A, Anterior; R, right.

Figure 6.

Multivariate Granger causality relationships among selected ROIs (for details of selection, see Results) in better (top) and poorer (bottom) performers. Relative strength of path weights (in arbitrary units) is indicated by a pseudocolor code. The actual path weights are tabulated in Table 2. L, Left; R, right.

Figure 7.

Paths differing significantly between better and poorer groups on multivariate Granger causality analyses. Top, Better > poorer; bottom, poorer > better. L, Left; R, right.