Abstract
Severe Hemispatial Neglect is predominantly observed with right hemisphere (RH) parieto-temporal lobe damage and rarely observed with left hemisphere (LH) damage. The Representational Model of Neglect (Heilman and Van Den Abell, 1980) accounts for this asymmetry by suggesting that RH codes space bilaterally, while LH codes only contralateral space; however visual mapping of the parietal lobe fails to support this model (e.g., Swisher et al., 2007). Recently, our lab (Sheremata et al., 2010) observed that a hemispheric asymmetry emerged within visuotopically mapped parietal regions as visual short-term memory (VSTM) load increased; visuotopic RH parietal structures coded bilateral VSTM targets, while LH parietal structures coded contralateral targets. Curiously, no behavioral differences were observed with unilateral target presentation. To account for these findings, we propose a Dynamic Representational Model: 1. RH capacity > LH capacity; 2. hemispheres normally code contralateral space; 3. RH, but not LH, shifts resources to the ipsilateral field if contralateral load is low. The switching property of this model accounts for equal hemifield performance with unilateral targets, and makes the novel prediction that if targets are spread bilaterally, VSTM capacity in the LVF should exceed capacity in RVF, because the RH would be less able to aid the LH in RVF. We tested this prediction in behavioral change-detection VSTM experiments (N=40), using bilateral and unilateral target conditions. With unilateral presentation, LVF and RVF performance was equal; however, with bilateral presentation, capacity was significantly lower in the RVF (p <0.05). These findings confirm model predictions. A second experiment (N=40) investigated influences of bilateral distractors and failed to observe an asymmetry (p>.5). We conclude that the presence of LVF targets, but not distractors, occupies the RH, limiting its ability to aid the LH in coding RVF targets. These experiments support the view that RH spatial coding changes dynamically with task demands.
Meeting abstract presented at VSS 2012