Neglect's perspective on the Ponzo illusion.

Visual illusions have been used to explore implicit perception in neglect. Previous studies have highlighted differences between length and surface illusion perception in neglect, but much less is known about depth illusion perception. In the Ponzo illusion (a classic depth illusion), two converging oblique lines modulate the perceived length of two horizontal lines. In the current study, we presented modified versions of the Ponzo illusion in which only one of the converging oblique lines was presented (alternatively the right or the left one). This manipulation allowed us to explore (1) how acute patients with neglect process depth illusions, and (2) whether awareness of both converging lines is necessary for the full effect of the illusion. To examine these questions, we had participants (i.e. healthy controls, patients with neglect and right brain-damaged patients) to make a perceptual judgment regarding the perceived length of the upper versus lower horizontal line within the Ponzo frame in four conditions: (1) the classic Ponzo illusion, (2) a "modified left" Ponzo illusion with a single oblique line on the left, (3) a "modified right" Ponzo illusion with a single oblique line on the right and (4) a control condition with parallel lines. The results indicated that all participants perceived the canonical Ponzo illusion and the modified right illusion. Critically, patients with neglect did not perceive the modified left illusion. In addition, for neglect patients, there was no difference in the strength of the perceived illusion when comparing the canonical illusion with the modified right illusion. Importantly, single case analysis revealed a high degree of variability in the neglect group that seemed to be linked with the amount of damage to occipital areas. Overall our results indicate that: (1) the classic Ponzo illusion might be perceived in neglect patients based solely on perception of the right side of the stimulus configuration, and (2) differences between types of illusions (i.e. depth vs. length), and variability between patients suggest that caution is needed when utilizing these kinds of illusions to assess implicit processing in neglect.