Ocular and Image Components in Binocular Rivalry: Measuring their strengths and decay rates

摘要

In binocular rivalry, when different, incompatible images are presented to the two eyes and only one is perceived, there are at least two factors that determine which image will dominate in perception: (1) a competition between the two eyes, with a dominant eye tending to determine perception independent of what image it is receiving, and (2) an image competition in which an image that is currently perceived tends to continue being perceived independent of which eye is receiving it. Here we demonstrate a paradigm for measuring the instantaneous a??strengtha?? of each of these components (ocular and image) and how these strengths decay over time. The experiments were carried out by Bartels and Logothetis (unpublished) who presented different, incompatible flower images to the left and right eyes. Usually, only one of the two images was perceived (binocular rivalry). After either 0.3 or 3.0 sec, there was a brief interruption stimuus, subsequently either the same stimuli were presented again or the stimuli were exchanged between eyes. The probability of perceiving a particular image after the interruption is represented in a model that assumes (1) the instantaneous ocular strength of the eye that is receiving the dominant image is represented by a positive number, and the suppressed eye by its negative; a second number represents the strength of the dominant image, its negative represents the suppressed image; (2) ocular- and image-component strengths add linearly, (3) the summed strengths are perturbed by additive, Gaussian internal noise; (4) the highest-strength eye+image combination is perceived after the interruption. The transformation of the raw probability-of-perception data into strengths removes all interactions. For these images, ocular strength was greater than image strength. For all subjects individually for all conditions investigated, the strengths of ocular and image components decline over time at same rate, completely in parallel.