Purchase this article with an account.
Sandrine Chemla, Alexandre Reynaud, Guillaume Masson, Frederic Chavane; V1 population activity represents global motion velocity of long-range apparent motion in the awake monkey. Journal of Vision 2015;15(12):480. doi: https://doi.org/10.1167/15.12.480.
Download citation file:
© ARVO (1962-2015); The Authors (2016-present)
Two stationary stimuli successively flashed in spatially separated positions generates the so-called apparent motion illusion. The illusion depends on the precise spatial and temporal separations of the stimuli and is called long-range apparent motion (lrAM) for large spatiotemporal(ST) separations. Since these values extend well beyond the typical selectivity of early visual neurons, it is unclear how the visual system computes motion signals from such sequence of static stimuli. We investigated whether a global motion representation could emerge at the level of V1 population in response to a two-stroke lrAM through lateral interactions. We used voltage-sensitive dye imaging to study in real-time the V1 population activity of two fixating monkeys in response to lrAM stimulations, whose velocity(direction and speed) was manipulated. We observe the emergence of a ST representation of the global motion:a cortical correlate of the illusory motion. This ST representation of V1 population is shaped by non-linear interactions:the apparition of the second stimulus generates a spread of suppression in the direction opposite to the AM at a speed compatible with the horizontal propagation, independent of the stimulus' speed. Such spread of suppression acts as a fast normalization that optimally shapes the ST representation of global motion along the AM path, to accurately represent the stimulus velocity. To validate this hypothesis, we applied an opponent motion energy model on the observed and linearly-predicted V1 activity. The model systematically produced the largest energy for the appropriate direction and speed on the observed activity and failed for the linear prediction. Our results suggest that motion signal can emerge at the level of V1 population in response to lrAM, a signal that could then be decoded by downstream areas.  Cavanagh & Mather(1989). Spat.Vis.,4(2-3),2-3.  Muller et al.(2014).Nat.Comm.,5.  Chemla & Chavane(2010).J.Physiol-Paris,104(1),40-50.  Jancke et al.(2004).Nature,428:423-6.  Reynaud et al.(2012).J.Neurosci,32(36),12558-12569.  Adelson & Bergen (1985).JOpt.Soc.Am.A,2:284-299.
Meeting abstract presented at VSS 2015
This PDF is available to Subscribers Only