Neural correlates of disambiguating an ambiguous multisensory motion stimulus
Benjamin Taylor Files, Lynne E Bernstein
Poster
Time: 2009-06-29 11:00 AM – 12:30 PM
Last modified: 2009-06-04
Abstract
Vision appears to be the dominant influence when we interpret motion in the world around us, but information from the other senses can help to interpret an otherwise ambiguous stimulus. An example of this is the bounce/stream illusion. The bounce/stream illusion arises in the context of an ambiguous visual display in which two identical objects approach each other, overlap and then move apart. This visual display is typically interpreted in one of two ways: the two objects exchanged trajectories at the point of overlap (i.e., they ‘bounced’ off each other) or the two objects each continued on their original trajectories (i.e., they ‘streamed’ through each other). Although there is no information in the display itself to bias an observer toward one interpretation or the other, in a majority of trials the stream response is reported. However, when an auditory stimulus is added to the presentation at the moment of overlap, this tendency reverses and the bounce response becomes dominant.
We sought to determine if the bounce/stream illusion is due to a perceptual or post-perceptual effect. 64-channel EEG plus two bipolar eye-movement channels were used to record continuous EEG while participants viewed visual-only (VO) and audio-visual (AV) bounce/stream displays plus an additional audio-only control. All the stimuli were in pseudo-randomized order with jittered inter-stimulus intervals. Continuous EEG data were epoched around stimulus onset and sorted by condition (AV or VO) and response (Bounce or No Bounce). In the visual-only condition, a response-related difference between the averaged ERPs emerged approximately 100 ms after the overlap. In the audio-visual condition, a response-related difference emerged approximately 200 ms after the overlap. Differences with these latencies are consistent with a bottom-up perceptually-driven effect. However, within the first 75 ms after the visual stimulus onset (approximately 525 ms before the moving objects overlapped), a difference on occipital and parieto-occipital electrodes emerges across the two response types for both VO and AV stimuli. This differential response well before the actual overlap event suggests that the early visual response indicates a state of the neural networks involved in eventually disambiguating this ambiguous event. In this presentation, we will describe coherence differences across conditions and response types. The early effects suggest that the pre-perceptual brain state is critically responsible for the perceptual illusion. (Research supported by NIH/NIDCD DC008308.)
We sought to determine if the bounce/stream illusion is due to a perceptual or post-perceptual effect. 64-channel EEG plus two bipolar eye-movement channels were used to record continuous EEG while participants viewed visual-only (VO) and audio-visual (AV) bounce/stream displays plus an additional audio-only control. All the stimuli were in pseudo-randomized order with jittered inter-stimulus intervals. Continuous EEG data were epoched around stimulus onset and sorted by condition (AV or VO) and response (Bounce or No Bounce). In the visual-only condition, a response-related difference between the averaged ERPs emerged approximately 100 ms after the overlap. In the audio-visual condition, a response-related difference emerged approximately 200 ms after the overlap. Differences with these latencies are consistent with a bottom-up perceptually-driven effect. However, within the first 75 ms after the visual stimulus onset (approximately 525 ms before the moving objects overlapped), a difference on occipital and parieto-occipital electrodes emerges across the two response types for both VO and AV stimuli. This differential response well before the actual overlap event suggests that the early visual response indicates a state of the neural networks involved in eventually disambiguating this ambiguous event. In this presentation, we will describe coherence differences across conditions and response types. The early effects suggest that the pre-perceptual brain state is critically responsible for the perceptual illusion. (Research supported by NIH/NIDCD DC008308.)