Visual attention with implications for unilateral spatial neglect.

Abstract

Recent models of visual attention (eg. Rizzolatti et al., 1987) have suggested that a similar system orients visual attention as is used to produce a saccadic eye movement. This thesis provides further support for the link between the attentional and eye orienting systems and has incorporated ideas from recent models of saccade generation. The time taken by normal subjects to initiate a saccade ('latency'), is examined in Chapters two, three and four. Subjects were given attentional instructions and saccades made to either: unilateral single, or, bilateral double, targets. Latency to attended targets was not greatly enhanced, while latency to non-attended targets was greatly slowed. The results support both the premotor model of visual attention and models of visual attention that emphasise the inhibitory consequence of directed attention. Bilateral double targets produced an additional slowing on saccade latency, which could reflect a further automatic attentional inhibition produced in the contralateral field by the stimulus onset. Fixation point offset (in 'gap' situations) is known to reduce saccade latency, which has been attributed to prior attentional disengagement (Fischer, 1987). In Chapter two, the use of a gap situation produced a generalised speeding which was independent of the effects of directed visual attention. This suggests that active fixation affects a separate component to that involved in orienting visual attention to a spatial location. This idea is incorporated into a model which emphasises the inhibitory consequences of attentive fixation. Chapters six and seven report the findings from an experimental investigation of a patient (B.Q.) with a 'unilateral spatial neglect', a condition often attributed to a deficit of visual attention. The 'gap' paradigm was shown to be effective at reducing the severity of B.Q.'s contralateral neglect. In contrast to normal subjects, bilateral double targets did not have an inhibitory effect on her saccade latency. These findings are explained in terms of a model that neglect results in part from the loss of attentional inhibition for the ipsilesional side of space and in part an inability to switch off contralesional inhibition produced during active fixation. A functional model is proposed in Chapter eight to account for the findings. This supports the close link between the attentional orienting and saccade programming systems. An additional implication of the findings is that models of visual attention and saccade generation need to consider the inhibitory consequences of directing attention to a spatial location.