Abstract
We investigated the coordination of saccade target selection and saccade production by measuring saccade endpoint variability with and without visual distractors. Four macaque monkeys performed both efficient pop-out (red/green) and inefficient form (T/L) visual search task over 299 sessions. Each trial consisted of 2, 4, or 8 iso-eccentric items, one of which was the target. We compared the distributions of variability in polar angle and amplitude of endpoints of saccades directed to the target with and without distractors. We found that visual search distractors influence saccade production. Irrespective of search efficiency, monkeys demonstrated increasing variability of endpoints in polar angle with increasing search set size. We interpret this finding in relation to the pattern of presaccadic activation in superior colliculus and frontal eye field motor maps. The center of gravity of the pattern of activation in the motor map dictates saccade endpoints (van Optal & van Gisbergen, 1989 Vision Res. Vol 29:1183). Greater variation in endpoints entails greater variation in this center of gravity, implying that neural saccade target selection does not entirely filter out distractors. The magnitude of systematic and random scatter of saccade endpoints during visual search is important to characterize because ideal observer models of visual search behavior assume that saccade endpoints are an accurate readout of visual processing (e.g. Beutter et al. 2003 J Opt Soc Am A Opt Image Sci Vis. 20:1341; Najemnik & Geisler 2005 Nature. 434:387).
Meeting abstract presented at VSS 2014