Figure 1. Paradigm, behavioural and lesion data. Subjects performed a masked prime task (A). Arrow stimuli subtended approximately 1.5×1°. Neutral primes comprised the arrows rearranged forming a square that carried no directional information (not shown). Twelve blocks of 24 stimuli contained six different trial types randomised with the constraint that each condition occurred the same number of times per block. Hands were covered during the experiment to prevent visual guidance of movement. Subjects were instructed to fixate the laptop display centrally and eye position was monitored by the experimenter. A practice session (<2 min) took place beforehand. Only patients with motor neglect showed significant reaction time delay when a right prime preceded a movement with the left hand (B, red circle). Hence right hand motor plans significantly intrude on left hand movement, but not vice versa, only in patients with motor neglect. Lesions were plotted using MRICro software (www.mricro.com) from either CT or MR. Lesion subtraction (patients with motor neglect minus those without) shows frontal white matter selectively affected in patients with motor neglect (C). (D) Brunner–Munzel statistic revealed that right putamen and subcortical white matter were significantly associated (z scores >4.47) with abnormal performance in the masked prime task using the left, but not the right, hand. (E) Severity of motor neglect is significantly associated (z >4.49) with damage at several discrete areas within the right hemisphere (including white matter near the putamen, inferior frontal gyrus, rolandic operculum and parietal supramarginal gyrus). RT, reaction time; SOA, stimulus onset asynchrony.