Purchase this article with an account.
James S. McDonald, Zoe Kourtzi; Perception of motion induction for naturalistic images in the human visual cortex. Journal of Vision 2005;5(8):668. doi: 10.1167/5.8.668.
Download citation file:
© ARVO (1962-2015); The Authors (2016-present)
Our visual perception often differs from the physical reality of the natural world. We exploited this discrepancy, with a perceptual illusion known as motion induction, to examine motion processing of naturalistic stimuli in the human visual cortex. When an incoherently moving stimulus is placed in a coherently moving surround, observers perceive the incoherent pattern moving in the opposite direction to the surround. We used psychophysics and fMRI to investigate the neural basis of motion induction for naturalistic images. Specifically, we used natural texture stimuli with 1/f amplitude spectra that consisted of a central region in a moving surround. The surround moved up or down with 100% coherent motion. We manipulated the coherence of the motion of the center: the center moved (up or down) at different coherence levels between random and 100%. When tested psychophysically, subjects misreported the direction of motion of the center. When the center had 0% coherence, observers perceived it moving in the opposite direction to the surround. The observers' point of subjective equality (PSE), i.e. when observers reported the center moving up and down an equal number of times, occurred when it had 30–50% coherent motion in the same direction as the surround. Based on the physical properties of the stimulus, we predicted that fMRI responses would be lower at 0% coherence than at the PSE where the motion coherence is higher. Alternatively the perceptual results predicted that the difference between these two conditions would be absent, or even reversed. That is; stronger fMRI responses would be observed at 0% coherence where there is more perceived motion than at the PSE. Our results in hMT+/V5 showed that fMRI responses correlate with the perceptual rather than the physical coherence in the stimulus. Our findings suggest that the motion induction effect can be mediated by a combination of motion coherence and motion contrast mechanisms in hMT+/V5.
This PDF is available to Subscribers Only