Audio-visual interactions for motion perception in depth modulate activity in visual area V3A.

Multisensory signals can enhance the spatial perception of objects and events in the environment. Changes of visual size and auditory intensity provide us with the main cues about motion direction in depth. However, frequency changes in audition and binocular disparity in vision also contribute to the perception of motion in depth. Here, we presented subjects with several combinations of auditory and visual depth-cues to investigate multisensory interactions during processing of motion in depth. The task was to discriminate the direction of auditory motion in depth according to increasing or decreasing intensity. Rising or falling auditory frequency provided an additional within-audition cue that matched or did not match the intensity change (i.e. intensity-frequency (IF) "matched vs. unmatched" conditions). In two-thirds of the trials, a task-irrelevant visual stimulus moved either in the same or opposite direction of the auditory target, leading to audio-visual "congruent vs. incongruent" between-modalities depth-cues. Furthermore, these conditions were presented either with or without binocular disparity. Behavioral data showed that the best performance was observed in the audio-visual congruent condition with IF matched. Brain imaging results revealed maximal response in visual area V3A when all cues provided congruent and reliable depth information (i.e. audio-visual congruent, IF-matched condition including disparity cues). Analyses of effective connectivity revealed increased coupling from auditory cortex to V3A specifically in audio-visual congruent trials. We conclude that within- and between-modalities cues jointly contribute to the processing of motion direction in depth, and that they do so via dynamic changes of connectivity between visual and auditory cortices.