Visual perturbation paradigms in goal-directed reaching investigated with functional MRI and subsequent transcranial magnetic stimulation

Introduction Involvement of the left posterior parietal cortex (lPPC) in online motor control has been demonstrated in recent years using fMRI (Culham et al, 2006). However, the human homologue to the macaque parietal reach region, or even more detailed functional anatomy of processes involved in motor control, is still controversial (Iacoboni, 2006). The main challenge is the spatial co-localization of functions that are also involved in motor execution, e.g. saccades, and motor planning (Astafiev et al, 2003). TMS offers the possibility to disentangle these functions due to its high temporal resolution, and can discriminate necessary from co-activated brain areas. Desmurget et al (1999) showed that online correction for reaching to an altered target can be disturbed using TMS over the lPPC. Here, we test sub-regions of the lPPC for necessity in online correction to different visual perturbations. Methods Nine healthy, right-handed participants performed closed-loop (i.e., with visual feedback of the hand) reach-to-target tasks with different perturbation paradigms: Displacement of the visual target or displacement of the visual hand feedback, which allowed us to investigate the processes involved in body's effectors representation and the processes involved in representation of the environment, respectively. First, the participants were tested with an fMRI localizer task to assess putative lPPC sub-regions that are involved in online motor control. The fMRI localizer consisted of blocks for fixation, saccades, and reaching with an MR-compatible joystick. Within the reaching blocks, the visual perturbations (including none) were randomized in a fast event-related design. The different perturbation conditions were contrasted against the unperturbed reaching to assess activation related to online-correction, masked with general reaching activation. Maxima from the group analyses, individual's own maxima, and control sites were used as stimulation sites for subsequent event-related TMS studies. The TMS experiments were conducted in a VR environment with a robot arm to enable naturalistic but highly controllable conditions. Results On the group level, we found different peak fMRI activations in the lPPC for different visual perturbations (Figure 1). Additionally, most subjects had strong individual peak fMRI activations on sites without group activation (Figure 2). The perturbing effect of TMS for corrections to a visual target perturbation is correlated with the strength of the fMRI activations (Figure 3), with effects on sites of group fMRI activation, additional effects on sites of individual's fMRI activation, but none on control sites (Figure 4). The perturbing effect of TMS for corrections to a visual hand perturbation is concentrated at the single group fMRI maximum. Conclusions Widespread areas in the lPPC are crucial for processing of visual target information needed for online control of movements. These areas include SMG in addition to anterior IPS and SPL. The area necessary for visual hand information processing is a part of the areas needed for visual target processing. Planning TMS stimulation sites based on individual fMRI activations is a more successful approach than planning them based on group fMRI activation, which is still better than planning without (f)MRI. The TMS effect is spatially selective, thus demonstrating a good spatial resolution of the method