ausblenden:
Schlagwörter:
-
Zusammenfassung:
Primates are social animals whose communication is based on their conspecics' vocalizations and facial expressions. Although a lot of work to date has studied the unimodal
representation of vocal or facial information, little is known about the way the nervous
system supports the processing of communication signals from dierent sensory modalities
to combine them into a coherent audiovisual percept. It is thought that the brains
of human and nonhuman primates evaluate vocal expressions and facial information separately
in specialized 'voice' and 'face' brain regions but we wondered if cross sensory
interactions were already evident at the neuronal level in these typically unimodal brain
regions.
Using movies of vocalizing humans and monkeys as stimuli, we recorded extracellularly
from the auditory cortex of a macaque monkey, targeting his 'voice' region in the right
hemisphere. Within a multi factorial design we evaluated how these auditory neurons responded
to dierent sensory modalities (auditory or visual) or combinations of modalities
(audiovisual). We also analyzed the responses for species specic eects (human/ monkey
speaker), call type specicity (coo/ grunt), as well as speaker familiarity, size and identity.
Following the approach in the original fMRI study localizing the monkey voice region,
our recordings identied a voice area 'cluster' in this animal. Within this auditory cluster
of sites, we observed a signicant visual in
uence on both the local eld potential (LFP)
and the spiking activity (AMUA), and found that 30 of the sites showed audiovisual
interactions in the LFP signals, and 38 in the AMUA. Grunts were especially eective
stimuli for this region and rather than a specialization for monkey vocalizations, human
vocalizations also elicited strong responses.
Our results provide evidence for visual in
uences in what has been characterized as an
auditory 'voice' area suggesting that at least the 'voice' regions are in
uenced by the
visual modality. Voices and faces seem to already interact at traditionally unisensory
brain areas, rather than cross sensory information being combined only in higher-level,
associative or multisensory regions of the brain.
