Abstract
During the opponency stage of computational motion-processing models, the responses of two directionally-selective cells with opposing preferred directions are subtracted (Adelson & Bergen, 1985). This calculation outputs the local direction of motion and reduces further processing of non-informative flicker-noise. A biological implementation of opponency involving brain area MT is speculated to exist, since MT neurons are poorly driven by counter-phase motion stimuli. Counter-phase stimuli are created by spatially pairing oppositely-moving dots such that any local area contains balanced quantities of opposing signals (Qian & Andersen, 1994). If the weak MT response to counter-phase motion is truly indicative of a noise-reducing mechanism, then this response may resemble MT's response to flickery non-motion stimuli. Furthermore, when included as background distractors in a direction discrimination task, task-irrelevant counter-phase and non-motion stimuli may similarly affect behavioral performance. In the current project, we examined these predictions using psychophysics and fMRI. We created a non-suppressed in-phase stimulus to use as a control comparison by reversing the direction of one dot in each counter-phase pair. During the psychophysical experiment, participants judged whether target dots moved coherently leftward or rightward. This task was embedded within counter-phase, in-phase, and non-motion backgrounds. During the fMRI experiment, counter-phase, in-phase, and non-motion stimuli were presented without target dots to collect clean patterns of activation. As predicted, in-phase trials elicited high BOLD responses and high behavioral thresholds, while counter-phase and non-motion trials elicited similarly low BOLD responses and behavioral thresholds. Moreover, a three-way MVPA classification of MT fMRI data found good classification of in-phase stimuli, but poor discrimination between counter-phase and non-motion stimuli. All together, these results suggest that counter-phase and non-motion stimuli are processed similarly, strengthening the idea that the weak MT response to counter-phase motion is a signature of the brain's noise-reduction mechanism.
Meeting abstract presented at VSS 2017