Abstract
Purpose :
Past experience of regularities in our visual environment facilitates visual search. When distractor configurations are repeated, incidental learning leads to faster search times than search for the same target in random distractor layouts. This contextual cueing is severely reduced when central vision is lost, both in patients with AMD (Geringswald et al., J. Vis., 2013) and with gaze-contingent scotoma simulation, due to high visual working memory demands of visual exploration with a scotoma (Geringswald and Pollmann, JEP:HPP, 2015). Here, we investigated whether visual search training with a central scotoma simulation can improve contextual cueing in normal-sighted observers.
Methods :
We examined contextual cueing in five consecutive sessions, spread on average over 10 days, of a T-among-L visual search task (Figure 1, cf. Geringswald et al., Front. Hum. Neurosci. 2012) with novel and repeated search displays in young normal-sighted observers. Repeated displays were different in each session. We compared a gaze-contingent central scotoma simulation group (n = 23) to an unrestricted viewing control group (n = 24). The central scotoma was simulated as an opaque disk of 7° diameter, smoothly fading out at the border. The disk was of the same grey as the background, so that the black T and L stimuli disappeared into the background when covered by the scotoma.
Results :
In the control group, contextual cueing led to increasingly faster search in repeated than novel displays already in Session 1 (Figure 2), indicated by both a significant main effect of configuration (novel, repeated; F(1,23) = 6.21, p < .05) and a significant configuration x time (epochs 1-4) interaction (F(3,69) = 6.03, p < .05). In the scotoma group, contextual cueing was first observed in session 3 (configuration main effect: F(1,22) = 6.77, p < .05; interaction: F(3,66) = 4.38, p < .05).
Conclusions :
These results are in agreement with recent reports that visual search may adapt already after a few hours of search training to the presence of a simulated central scotoma (Kwon et al., Curr. Biol., 2013; Walsh and Liu, Vis. Res. 2014). We extend these findings by showing that such training, probably by reducing visuospatial working memory load, enables memory-guided search in the presence of central vision loss. Apart from its theoretical importance, this may show a promising way for training programs in patients with central vision loss.
This is an abstract that was submitted for the 2016 ARVO Annual Meeting, held in Seattle, Wash., May 1-5, 2016.