Neurology and Neurosurgery, McGill University, Montreal, Quebec, Canada.
McConnell Brain Imaging Centre, Montreal Neurological Institute, Montreal, Quebec, Canada.
PLoS One. 2019 Jul 11;14(7):e0218529. doi: 10.1371/journal.pone.0218529. eCollection 2019.
Binocular rivalry (BR) is a dynamic visual illusion that provides insight into the cortical mechanisms of visual awareness, stimulus selection, and object identification. When dissimilar binocular images cannot be fused, perception switches every few seconds between the left and right eye images. The speed at which individuals switch between alternatives is a stable, partially heritable trait. In order to isolate the monocular and binocular processes that determine the speed of rivalry, we presented stimuli tagged with a different flicker frequency in each eye and applied stimulus-phase locked MEG source imaging. We hypothesized that the strength of the evoked fundamental or intermodulation frequencies would vary when comparing Fast and Slow Switchers. Ten subjects reported perceptual alternations, with mean dominance durations between 1.2-4.0 sec. During BR, event-related monocular input in V1, and broadly in higher-tier ventral temporal cortex, waxed and waned with the periods of left or right eye dominance/suppression. In addition, we show that Slow Switchers produce greater evoked intermodulation frequency responses in a cortical network composed of V1, lateral occipital, posterior STS, retrosplenial & superior parietal cortices. Importantly, these dominance durations were not predictable from the brain responses to either of the fundamental tagging frequencies in isolation, nor from any responses to a pattern rivalry control condition, or a non-rivalrous control. The novel cortical network isolated, which overlaps with the default-mode network, may contain neurons that compute the level of endogenous monocular difference, and monitor accumulation of this conflict over extended periods of time. These findings are the first to relate the speed of rivalry across observers to the 'efficient coding' theory of computing binocular differences that may apply to binocular vision generally.
双眼竞争（BR）是一种动态的视觉错觉，它为皮质机制提供了洞察视觉意识、刺激选择和物体识别。当不同的双眼图像不能融合时，感知每几秒钟就在左眼和右眼图像之间切换。个体在替代物之间切换的速度是一种稳定的、部分可遗传的特征。为了分离决定竞争速度的单眼和双眼过程，我们在每只眼睛中呈现带有不同闪烁频率的刺激，并应用刺激相位锁定的 MEG 源成像。我们假设，当比较快切换者和慢切换者时，诱发的基本或互调频率的强度会有所不同。10 名受试者报告了知觉交替，平均优势持续时间在 1.2-4.0 秒之间。在 BR 期间，V1 中的事件相关单眼输入，以及在更高级的腹侧颞叶皮质中广泛地，随着左眼或右眼优势/抑制的周期而起伏。此外，我们还表明，慢切换者在由 V1、外侧枕叶、后 STS、后扣带回和顶叶上回组成的皮质网络中产生更大的诱发互调频率响应。重要的是，这些优势持续时间不能从大脑对孤立的基本标记频率中的任何一个的反应，也不能从任何对图案竞争控制条件的反应，或非竞争控制的反应中预测出来。分离出的新皮质网络与默认模式网络重叠，其中可能包含计算内源性单眼差异水平的神经元，并监测这种冲突在较长时间内的积累。这些发现是第一个将观察者之间竞争速度与可能适用于一般 binocular 视觉的“有效编码”理论联系起来的发现。
