Multisensory object-perception: Audio-visual binding studied with fMRI
Poster Presentation

Matthias Bischoff
Bender Institute of Neuroimaging, University Giessen

*Gebhard Sammer
Bender Institute of Neuroimaging, University Giessen

*Carlo Raimund Blecker
Bender Institute of Neuroimaging, University Giessen

*Bertram Walter
Bender Institute of Neuroimaging, University Giessen

*Dieter Vaitl
Bender Institute of Neuroimaging, University Giessen

     Abstract ID Number: 140
     Full text: PDF
     Last modified: July 18, 2005

Abstract
Brain-physiological correlates of audio-visual object perception are examined under the aspect of crossmodal integration. fMRI is recorded in order to reveal activation in areas discussed to be involved in multisensory processing in the literature.
Aiming at consciousness-related binding as subset of crossmodal integration, the study focuses on the comparison of a ''bound'' object with ''unbound'' object-features. Simple visual and auditory stimuli were presented simultaneously or with a time-delay. Simultaneous auditory-visual stimulation biases the perceived localization of the auditory stimulus towards the visual one. This is known as ventriloquism-effect. Trials, in which participants reported a ventriloquism-effect, were considered to represent consciousness-related object-binding. Asynchronously presented stimuli (no localization bias) refer to unbound object-features. Those trials formed the control (no-binding) condition.
The contrast of binding vs no-binding condition (N=19 subjects) revealed activation in the insula, superior temporal sulcus and parieto-occipital sulcus (ROI-analysis, FWE-corrected at p = 0.05).
Concluding, consciousness-related binding by synchronous audio-visual stimulation activates cortical areas, which have been suggested to serve as heteromodal areas in the recent literature. In contrast, these areas showed no activation in trials without a ventriloquism-effect (no-binding).