Spatio-temporal frequency characteristics of intersensory components in audiovisually evoked potentials

Perception of the external world is based on complex neural processes allowing for combination of sensory experiences from different modalities. Audiovisual (AV) integration is discussed in this paper on the basis of the intersensory component (IC), which is the part of the multisensory-evoked potential that is not explained by linear summation of the unisensory-evoked potentials. It was predicted that audiovisual ICs can be extracted, localized, and differentiated by means of wavelet-based frequency analysis. Healthy, right-handed subjects (n = 15) were instructed to view and listen to presented stimuli (A: auditory; V: visual; and AV: audiovisual). Electroencephalographic data was analyzed off-line by means of wavelet transformation utilizing quadratic B-spline mother wavelets. Cross-modal interaction was investigated by subtracting wavelet responses to unimodal stimuli (A, V) alone from the wavelet responses to the combined stimuli (AV; i.e., interaction = AV - (A + V)). These difference waveforms revealed the phase-locked fraction of ICs further characterized by frequency-band and location. Spatio-temporally distinct ICs were observed in all frequency bands [31-62 Hz (gamma), 16-31 Hz (beta), 8-16 Hz (alpha), 4-8 Hz (theta), 0.5-4 Hz band (delta)]. These were most pronounced and sustained in the theta frequency band with early (<100 ms) appearance in fronto-centro-parietal sites. In contrast, alpha-band ICs tended to appear later (>200 ms) in these locations. High-frequency (beta- and gamma-band) ICs were less organized in their spatial pattern with both early and late appearance. ICs may reflect sensory and cognitive/integrative processes at the cortical level. In case of intersensory processing, alpha- and theta-activity appear to be spatio-temporally distinct, and could therefore participate in different stages of perception. These findings add further support to current model views of oscillatory activity in selectively distributed networks.