A Microsaccadic Account of Attentional Capture and Inhibition of Return in Posner Cueing.

Xiaoguang Tian,Ziad M Hafed,Masatoshi Yoshida

doi:10.3389/fnsys.2016.00023

Abstract

Microsaccades exhibit systematic oscillations in direction after spatial cueing, and these oscillations correlate with facilitatory and inhibitory changes in behavioral performance in the same tasks. However, independent of cueing, facilitatory and inhibitory changes in visual sensitivity also arise pre-microsaccadically. Given such pre-microsaccadic modulation, an imperative question to ask becomes: how much of task performance in spatial cueing may be attributable to these peri-movement changes in visual sensitivity? To investigate this question, we adopted a theoretical approach. We developed a minimalist model in which: (1) microsaccades are repetitively generated using a rise-to-threshold mechanism, and (2) pre-microsaccadic target onset is associated with direction-dependent modulation of visual sensitivity, as found experimentally. We asked whether such a model alone is sufficient to account for performance dynamics in spatial cueing. Our model not only explained fine-scale microsaccade frequency and direction modulations after spatial cueing, but it also generated classic facilitatory (i.e., attentional capture) and inhibitory [i.e., inhibition of return (IOR)] effects of the cue on behavioral performance. According to the model, cues reflexively reset the oculomotor system, which unmasks oscillatory processes underlying microsaccade generation; once these oscillatory processes are unmasked, “attentional capture” and “IOR” become direct outcomes of pre-microsaccadic enhancement or suppression, respectively. Interestingly, our model predicted that facilitatory and inhibitory effects on behavior should appear as a function of target onset relative to microsaccades even without prior cues. We experimentally validated this prediction for both saccadic and manual responses. We also established a potential causal mechanism for the microsaccadic oscillatory processes hypothesized by our model. We used retinal-image stabilization to experimentally control instantaneous foveal motor error during the presentation of peripheral cues, and we found that post-cue microsaccadic oscillations were severely disrupted. This suggests that microsaccades in spatial cueing tasks reflect active oculomotor correction of foveal motor error, rather than presumed oscillatory covert attentional processes. Taken together, our results demonstrate that peri-microsaccadic changes in vision can go a long way in accounting for some classic behavioral phenomena.

Highlights

Microsaccades are small saccades that occur repeatedly during prolonged fixation (Hafed, 2011; Hafed et al, 2015)
Because microsaccades inescapably occur in a variety of experiments enforcing fixation, one wonders how large of a contribution these microsaccade-related changes in vision have in such experiments? It could be the case that these movements are rare enough to be completely inconsequential, or it could be the case that active peri-microsaccadic changes in vision play a substantial role, despite the diminutive size of the eye movements, much like active peri-saccadic changes in vision can dramatically alter the state of the visual system (Duhamel et al, 1992; Cai et al, 1997; Ross et al, 1997, 2001; Lappe et al, 2000; Tolias et al, 2001; Sommer and Wurtz, 2006; Pola, 2011; Morris et al, 2012, 2013; Zirnsak et al, 2014)
For short cue-onset-to-targetonset asynchronies (CTOA’s), subjects oriented to the target faster if it appeared at the cued location than if it appeared at the opposite location (Figure 3A, 47 ms CTOA, p = 1.1∗10−4, two-sided t-test between same and opposite)

Summary

Introduction

Microsaccades are small saccades that occur repeatedly during prolonged fixation (Hafed, 2011; Hafed et al, 2015). We chose as a case study the Posner cueing task, which had been used to great effect in advancing our understanding of covert visual attention (Posner, 1980). In this task, spatial cueing facilitates stimulus detection at the cued location relative to other locations (Posner, 1980). This latter phenomenon, termed “inhibition of return” (IOR; Posner and Cohen, 1984; Posner et al, 1985; Klein, 2000; Lupianez et al, 2006), was thought to reflect a cognitive mechanism through which “the nervous system favors novel information over information previously presented” (Posner et al, 1985)

Methods

Results

Conclusion