Vocal processing networks in the human and marmoset brain

Abstract

Understanding the brain circuitry involved in vocal processing across species is crucial for unraveling the evolutionary roots of human communication. While previous research has pinpointed voice-sensitive regions in primates, direct cross-species comparisons using standardized protocols are limited. This study utilizes ultra-high field fMRI to explore vocal processing mechanisms in humans and marmosets. By employing voice-sensitive regions of interest (ROIs) identified via auditory localizers, we analyzed response time courses to species-specific vocalizations and non-vocal sounds using a dynamic auditory-stimulation paradigm. This approach gradually introduced sounds into white noise over 33 seconds. Results revealed that both species have responsive areas in the temporal, frontal, and cingulate cortices, with a distinct preference for vocalizations. Significant differences were found in the response time courses between vocal and non-vocal sounds, with humans displaying faster responses to vocalizations than marmosets. We also identified a shared antero-ventral auditory pathway in both species for vocal processing, originating from the superior temporal gyrus. Conversely, a posterior-dorsal pathway was more prominent in humans, whereas in marmosets, this pathway processed both sound types similarly. This comparative study sheds light on both conserved and divergent auditory pathways in primates, providing new insights into conspecific vocalization processing.