Purchase this article with an account.
Javier Garcia, John Pyles, Emily Grossman; Neural mechanisms underlying motion opponency in hMT+. Journal of Vision 2007;7(9):396. doi: https://doi.org/10.1167/7.9.396.
Download citation file:
© ARVO (1962-2015); The Authors (2016-present)
Background. Previous psychophysical and physiological research has shown the direction of visual motion to be perceptually cancelled in displays with locally balanced opposing motion signals (Qian et al.,1994). The perceptual consequences of these specialized circumstances are thought to be a result of exploiting the center-surround antagonistic relationship of directionally-selective inputs in motion-selective brain area MT. Motion antagonism, which in unbalanced conditions signals transparency, has been associated with “local” neurons in MT (Born & Tootell,1992), is dependent on stimulus contrast (Pack et al., 2005), implicated in perception of object motion (Tadin et al., 2003), and may underlie direction discrimination learning (Lu et al., 2004). The following experiments investigate the interaction of stimulus size and contrast in displays with locally unbalanced and balanced motion. A second set of experiments measures the BOLD response in V1 and hMT+ as a function of these same parameters. Method. Limited lifetime random dot cinematograms were constructed with dot-pairs that parametrically vary in spatial proximity between the paired dots. Subjects make a 2AFC angle discrimination on each 133msec trial (clockwise or counter-clockwise relative to an unseen reference angle). Angular thresholds were collected as a function of contrast and stimulus aperture size. In a blocked fMRI experiment, BOLD responses to the paired random dot cinematograms were collected as a function of proximity of the dot-pairs, stimulus aperture size, and contrast. Results. Angular deviation thresholds increase (i.e. poorer performance) as dot-pair proximity decreases. Angular deviation thresholds improve for larger stimuli and higher contrasts. BOLD responses from hMT+ also vary as a function of dot proximity and contrast. Conclusions. Our results evidence locally antagonistic neural computations underlie direction discrimination, and reflect local summation of directionally-selective inputs (perhaps through “local” neurons in hMT+) which mediate perception of transparent motion with motion opponent paired-dot displays.
This PDF is available to Subscribers Only