Purchase this article with an account.
Thaddeus Czuba, Lawrence Cormack, Alexander Huk, Adam Kohn; Neuronal selectivity for directions of 3D motion in area MT. Journal of Vision 2013;13(9):608. doi: https://doi.org/10.1167/13.9.608.
Download citation file:
© ARVO (1962-2015); The Authors (2016-present)
The mechanisms of frontoparallel (2D) motion processing have been studied extensively and are relatively well understood. However, it remains unclear how the visual system encodes motion that includes a component toward or away from the observer (3D motion). Traditional accounts suggest that binocular 3D motion processing relies on changes in disparity over time, but recent work argues that in many situations the most relevant cue might be interocular velocity differences (IOVD): differences in horizontal velocity signals in the two eyes that are geometrically coincident with changes in disparity over time (Cumming & Parker, 1994; Czuba et al., 2010; Shioiri et al., 2000). We tested how the motion processing pathway encodes binocular 3D motion information by performing extracellular recordings in area MT of anesthetized macaques. We measured responses to a full matrix of monocular and binocular motion conditions using drifting sinusoidal gratings, varying absolute and relative temporal frequencies in the two eyes so that binocular conditions spanned a broad range of 3D motion trajectories. Many MT cells showed similar preferences for monocular motion in each eye and straightforward summation of these signals for binocular stimuli—these cells did not encode IOVD information. However, an interesting subset of cells exhibited robust IOVD information, evident either as opposite direction preferences for motion shown in each eye or strong nonlinear interactions for binocular motion. We also performed detailed measurements of disparity selectivity, and found cells selective for 3D directions of motion could be either sensitive or insensitive to static disparities. Together, our results suggest area MT contains robust signals for 3D motion processing through IOVD sensitivity. Our data also provide a promising framework for exploring how 2D and 3D directions of motion could be represented by a common population of neurons along known—classically 2D—motion pathways.
Meeting abstract presented at VSS 2013
This PDF is available to Subscribers Only