Neural mechanisms underlying reaching for remembered targets cued kinesthetically or visually in left or right hemispace

Abstract

Reaching for a target involves integrative coordinate transformation processes between the representation of the target location, the sensorimotor information of limb of reach, and body space. Although right hemisphere dominance for visuospatial information processing is well established, corresponding right hemisphere dominance for kinesthetic spatial information processing remains to be demonstrated. We explored neural mechanisms of encoding target locations using 15O-butanol positron emission tomography (PET) in normal volunteers in a factorial experiment, where modality (visual/kinesthetic) and hemispace of target presentation (left/right of midsagittal plane) were varied systematically. After target presentation, subjects reached to the encoded target location. PET data analysis using SPM99 showed increased neural activity (P < 0.05, corrected) associated with left hemispace target presentation in right hemisphere areas (sensorimotor, anterior cingulate, insular, and temporo-occipital cortex) only. By contrast, right hemispace target presentation activated bilateral temporo-occipital cortex, which extended into the right temporo-parietal cortex and left sensorimotor cortex. A significant interaction of hemispace and modality of target presentation observed in right temporo-parietal cortex resulted from an increase in neural activity with kinesthetic target presentation in right hemispace. The data support an important role for the right temporo-parietal area in visuospatial processing and suggest a specific role of the right hemisphere in kinesthetic spatial processing.

Figure 1

Experimental setup. Subject positioned supine in the PET gantry with arms freely movable. OPTOTRAK 3020 system was used to acquire 3‐D position data. The small red LED target was presented by a robot. Vision was regulated using electro‐optic shutter‐glasses (PLATO). The dotted line from shutter‐glasses to mirror represents subject's requirement to focus on a fixation cross. Computers, EOG electrodes, IREDs, and cabling have been omitted to enhance clarity. All targets were presented within subject's reach.

Figure 2

Experimental design. A: Targets were presented visually or kinesthetically (factor 1: mode of target presentation) on right or left of subject's midsagittal plane (factor 2: hemispace of target presentation). Visual conditions: targets were placed in a position within the subject's right (VR) or left (VL) hemispace while vision was occluded using a spectacle mounted electro‐optic shuttering device. Shutter‐glasses were opened and target location disclosed while subject maintained focus on a central fixation aid. After visual target presentation, shutters were closed and the target removed. In kinesthetic conditions, subject's vision was occluded at all times. Subject's right (KR) or left (KL) arm was guided to the target location and back to the starting position. After target presentation, in all conditions a tone cued subjects to reach to the remembered target location. Subjects were asked to place their fingertip such that it would touch the target if present, and instructed to hold this position for about one sec before returning to the starting position. Filled circles, target position; black cross, fixation point. Visual information regulation: open circles, full vision; closed circles, vision occluded. B: Relative timing of target presentation, movement execution, and target change for all conditions.

Figure 3

Relative rCBF increases associated with hemispace of target presentation. Left vs. right hemispace presentation (VL + KL > VR + KR) (top) relative to right vs. left hemispace presentation (VR + KR > VL + KL) (bottom). Areas of significant relative rCBF increases (P < 0.05, corrected for multiple comparisons; extent threshold 0 voxels) are shown as through‐projections onto representations in standard stereotactic space [Talairach and Tournoux,1988; Friston et al.,1995]. R, right; L, left; P, posterior; A, anterior.

Figure 4

Interaction of hemispace with mode of target presentation when pointing to remembered targets. Areas of significant relative rCBF increases (P < 0.05 corrected for multiple comparisons; extent threshold 0 voxels) are shown as through‐projections onto representations of standard stereotactic space [Talairach and Tournoux,1988] for sagittal and coronal views. Transverse SPM_z‐maps were superimposed on group mean MR images, which had been spatially normalized into the same anatomic space [Talairach and Tournoux,1988]. Exact local maxima coordinates within activation areas in right temporo‐parietal cortex and the respective Z‐statistic are given in Table I. For the region in right temporo‐parietal cortex (indicated by green open circles), analysis of associated parameter estimates undergoing significant activation reveals specific involvement of this region when targets are presented kinesthetically on the right. Adjusted means rCBF (to a mean of 50 ml/dl/min) ± SEM for each condition are displayed for local maxima. VR, visual presentation/right hemispace; VL, visual presentation/left hemispace; KR, kinesthetic presentation/right hemispace; KL, kinesthetic presentation/left hemispace; R, right; L, left; P, posterior; A, anterior.