Purchase this article with an account.
Yulia Lerner, Nava Rubin; Neural modulations during perceptual bi-stability away from equi-dominance are common to two different ambiguous displays. Journal of Vision 2009;9(8):279. doi: https://doi.org/10.1167/9.8.279.
Download citation file:
© ARVO (1962-2015); The Authors (2016-present)
Perceptual bi-stability occurs when a stimulus offers two distinct, plausible interpretations. Although recent studies significantly advanced our understanding of the neural basis of bi-stability, it remains unclear whether there is a central mechanism of perceptual decision that weights sensory signals for all types of stimuli, or whether switches originate from competition in stimulus-specific networks. To address this question we employed the observation that the prevalence of each competing interpretation - the fraction of time it is dominant - varies continuously with stimulus parameters. While previous fMRI studies used bi-stable stimuli where the two interpretations were roughly equi-dominant, we studied neural activity for parametric configurations away from, as well as at equi-dominance. We used two different types of ambiguous stimuli with correspondingly different, stimulus-relevant parameters: for plaids (Wallach, 1935), increasing the angel between gratings' motion directions increases the prevalence of the transparent-gratings percept; for the occluded diamond (Lorenceau & Shiffrar, 1992), increasing the contrast of the occluders increases the prevalence of the whole-diamond percept. Three different parametric values were used for each stimulus type, chosen so that one interpretation was dominant 75% of the time (prevalent), 50% (equi-dominant), and 25% (non-prevalent), in separate 5 minute runs. In visual cortex, modulations were tied to percept identity (e.g., more active during ‘coherency’), regardless of prevalence (i.e., both when ‘coherency’ was dominant 75% and when it was dominant 25% of the time). In contrast, more anterior regions showed modulations tied to prevalence, not percept - some more active when the prevalent percept was dominant (STS) while many others showed the opposite modulation pattern (pITS, IPS, PoCS, CS, IPrCS, LS). The prevalence-modulation maps were remarkably similar for the two stimulus types, supporting the hypothesis that, additionally to stimulus-specific modulation in sensory cortices, bi-stable alternations involve higher level, distributed cortical networks that are common to different ambiguous stimuli.
This PDF is available to Subscribers Only