Space- and object-based attention in patients with a single hemisphere following childhood resection

Abstract

The neural processes underlying attentional processing are typically lateralized in adults, with spatial attention associated with the right hemisphere (RH) and object-based attention with the left hemisphere (LH). Using a modified two-rectangle attention paradigm, we compared the lateralization profiles of individuals with childhood hemispherectomy (either LH or RH) and age-matched, typically developing controls. Although patients exhibited slower reaction times (RTs) compared to controls, both groups benefited from valid attentional cueing. However, patients experienced significantly higher costs for invalid trials-reflected by larger RT differences between validly and invalidly cued targets. This was true for invalid trials on both cued and uncued objects, probes of object- and space-based attentional processes, respectively. Notably, controls showed no significant RT cost differences between invalidly cued locations on cued versus uncued objects. By contrast, patients exhibited greater RT costs for targets on uncued versus cued objects, suggesting greater difficulty shifting attention across objects. We explore potential explanations for this group difference and the lack of difference between patients with LH or RH resection. These findings enhance our understanding of spatial and object-based attention in typical development and reveal how significant neural injury affects the development of attentional systems in the LH and RH.

Figure 1.. The two-rectangle attention paradigm of Egly, Driver, et al., 1994.

Two rectangles appear on the screen, followed by an ‘exogenous’ or bottom-up cue (black outline), in one of four locations. After a brief pause, usually 100–150 ms, the target (black square) appears at one end of one rectangle in one of three possible locations (never occurs on the diagonal opposite the cued corner). In Valid (V) trials, the target appears in the same location and object (rectangle) as the cue – this occurs on roughly 70% of the trials, leading the observer to use the cue to predict target location. In the Invalid Spatial (IS) trials (second column), following the cue, the subsequent target appears at a different location, at the same relative position on the other, uncued object. In the Invalid Object (IO) trials (third column), following the cue, the subsequent target appears at an uncued location that is spatially equidistant from the cue as in IS trials but is within the same cued object. Neutral (N) trials (fourth column), are the baseline condition in which all four ends of the rectangles are cued, offering no prediction of upcoming location of the target.

Figure 2.. Two-rectangle attention paradigm modified to accommodate hemianopia in hemispheric surgery patients.

Participants were instructed to fixate on ‘Doug the Dog’ and help him decide which colored ball (red or purple) appeared on his bones at any given time by pressing one key if they saw the red ball and another key if they saw the purple ball. They were not informed about the presence of the pre-cue that flashed in either a valid location (70% of trials), an invalid spatial location on the opposite object of the target to be presented (10% of trials, ‘IS’), an invalid location on the same object of the target to be presented (but equidistant to the cue to target distance of IS trials;10% of trials, ‘IO’), or in all locations (neutral trials: 10%). After the cue appeared and a jittered inter-stimulus interval (between 100–150 ms) passed, participants had 2000 ms to identify the color of the target. A happy ‘Doug’ was shown if a response was completed in the allotted time irrespective of accuracy, and a sad ‘Doug’ was shown if no response was completed.

Figure 3.

Distribution of (a) reaction times and (b) invalid cue costs for patients (red) and controls (blue) for each trial type and cost type, respectively. The line in each box represents the median, while the lower and upper ranges represent the 25^th and 75^th percentiles, respectively; the whiskers extend to 1.5 times the interquartile range. Reaction time costs were calculated for each participant by subtracting their median RT on valid trials from that on invalid object trials (invalidO cost) and their median RT on valid trials from that on invalid spatial trials (invalidS cost). The lines above the boxplots represent post-hoc comparisons that were tested between trial/cost types within controls (blue) or within patients (red) and between groups within cost type (black). Solid lines represent significant differences and dashed lines represent differences that were tested but were not significant. As stimuli were only presented to patients’ intact hemifield, i.e., the hemifield opposite to their intact hemisphere. Thus, LH surgery patients saw stimuli in their left hemifield (circles) and RH surgery patients saw stimuli in their right hemifield (triangles). Statistical details are included in the main text.