Ocular search during line bisection. The effects of hemi-neglect and hemianopia.

We examined ocular fixations during line bisection in five patients with left hemianopia, two patients with right hemianopia, nine patients with left hemi-neglect and nine normal control subjects. Compared with measures in control subjects, the median fixation, and left- and rightmost fixations were shifted contralaterally in patients with hemianopia alone and ipsilaterally in patients with hemi-neglect. The fixation with the longest duration and the bisection point were also shifted contralaterally with hemianopia and ipsilaterally with hemi-neglect. However, the number of fixations and the spatial range spanned by fixations did not differ between the groups, showing that ocular exploration was not truncated in any group. Only some patients showed a previously reported directional search bias. Overall, there was no directional bias in saccadic number or amplitude. The distribution of fixations was most dense at the centre of the line in normal subjects, while hemianopic patients fixated most frequently at the ends of lines in their contralateral (blind) hemispace and at a central locus that was biased slightly contralaterally, as was their bisection judgement. This contralateral bias may reflect either an adaptive contralateral attentional gradient or a non-veridical spatial representation within the remaining normal hemifield. Hemi-neglect patients had a broad distribution of fixation peaks in the ipsilateral hemispace. Of two hemi-neglect patients with many fixations, one clustered fixations at a position right of centre, as if a normal fixation pattern was shifted rightward, while the other had two fixation peaks: one to the far right and the other near the centre of the line, reminiscent of the dual peaks of activity seen in some recent hemi-neglect models. These data reveal a heterogeneity in the routes by which right-biased judgements of spatial centre are reached by hemi-neglect patients.