Item

Abstract

Many animals use cross-sensory information during communication, but it remains unclear how the brain integrates face and voice information. Functional imaging evidence suggests that the brains of human and nonhuman primates contain voice- and face-sensitive regions, and some of the human studies have suggested that multisensory interactions occur in these regions. Yet, to date neurons in monkey voice/face regions have been studied exclusively with unisensory stimuli. We targeted neurons in a recently identified voice-sensitive cluster in the right hemisphere on the supratemporal plane to investigate how neurons in the monkey brain combine auditory voice and visual face information. Extracellular recordings were conducted in two Rhesus macaques participating in a visual fixation task. Dynamic face and voice stimuli (movies of vocalizing monkeys and humans imitating monkey “coo” calls) were presented in auditory only, visual only and audio-visual stimulation conditions, including congruent and incongruent audio-visual pairs. In this region, we identified spiking activity driven by the presence of auditory stimuli (n = 130 single- and multi-units), 42 of which demonstrated visual influences. Most of the visual modulation (36 of responsive units) consisted of nonadditive multisensory effects, where the audiovisual responses significantly deviated from the sum of both unimodal responses. The magnitude of the visual influences was differentially sensitive to stimulus features such as call type, speaker identity and familiarity. Human voices elicited qualitatively similar auditory and audiovisual responses as monkey voices. Finally, we found that incongruent stimuli elicited a larger proportion of sublinear audiovisual interactions, relative to congruent audiovisual pairs. Our results identify visual influences at the neuronal level in a primate auditory 'voice' region. Together, with results from functional imaging studies in humans, these findings extend our understanding of the multisensory influences at voice regions, which might also be evident in neurons at face-sensitive regions.