Motion regions are modulated by scene content

Abstract

Creating a stable perception of the world during pursuit eye movements is one of the everyday roles of visual system. Some motion regions have been shown to differentiate between motion in the external world from that generated by eye movements. However, in most circumstances, perceptual stability is consistently related to content: the surrounding scene is typically stable. However, no prior study has examined to which extent motion responsive regions are modulated by scene content, and whether there is an interaction between content and motion response. In the present study we used a factorial design that has previously been shown to reveal regional involvement in integrating efference copies of eye-movements with retinal motion to mediate perceptual stability and encode real-world motion. We then added scene content as a third factor, which allowed us to examine to which extent real-motion, retinal motion, and static responses were modulated by meaningful scenes versus their Fourier scrambled counterpart. We found that motion responses in human motion responsive regions V3A, V6, V5+/MT+ and cingulate sulcus visual area (CSv) were all modulated by scene content. Depending on the region, these motion-content interactions differentially depended on whether motion was self-induced or not. V3A was the only motion responsive region that also showed responses to still scenes. Our results suggest that contrary to the two-pathway hypothesis, scene responses are not isolated to ventral regions, but also can be found in dorsal areas.