Neglect patients often show deficits in responding to targets in the contralesional side of space. Past studies were able to ameliorate these deficits through manipulation of visual input. Here, the neural bases of the recovery of space following virtual reality (VR) training in neglect patients were investigated. Neglect patients were trained to respond to targets in the left side of space that appeared in the central or in the right side of space in a VR system. It was found that only patients with lesions that spared the inferior parietal/superior temporal regions were able to benefit from the VR training. It was concluded that these regions play a crucial role in the recovery of space that underlies the improvement of neglect patients when trained with VR. The implications of these results for determining the neural bases of a higher order attentional and/or spatial representation and for treating patients with unilateral neglect are discussed.

Previous studies have implicated the human parietal lobes in the on-line guidance of action. However, no study to date has examined at what stage in the on-line adjustment process do the parietal lobes play their most critical role. Repetitive transcranial magnetic stimulation (rTMS) was applied over the left intraparietal sulcus as participants reached to grasp a small or large illuminated cylinder. On some trials, the illumination could suddenly switch from the small to large cylinder, or vice-versa. Small–Large switches were associated with relatively early grip aperture adjustments, whereas Large–Small switches were associated with relatively late grip aperture adjustments. When rTMS was applied early in the movement, it disrupted on-line adjustments to Small–Large target switches, but not to Large–Small switches. Conversely, when rTMS was applied late in the movement, it disrupted adjustments to Large–Small target switches but not to Small–Large switches. The timing of the disruption by rTMS appeared linked to the initiation of the adjustment. It was concluded that the left parietal lobe plays a critical role in initiating an on-line adjustment to a change in target size, but not in executing that adjustment. The implications of these results for current views of on-line control are discussed.