Fig. 1: Overview of fMRI Methodology and Experimental Design.

a Photograph of the 9.4 Tesla (T) ultra-high field MRI scanner used for functional imaging with details about specifications of the gradient strength and the multi-channel receive coil (left), and the custom MRI-compatible restraint system featuring an 8-channel receive coil, with a marmoset secured by a head post and equipped with MRI-compatible auditory tubes (right). All elements were photographed by the authors. b Representation of the temporal signal-to-noise ratio (tSNR) achieved at 9.4 T, indicating the high-quality data acquisition enabled (adapted from⁶¹ with permission). c Schematic of brain coverage, designed by the authors, illustrating the fMRI scan achieved through 42 axial slices with an isotropic resolution of 0.5 mm, superimposed on the NIH marmoset brain template¹⁰⁴ (publicly available). d Depiction of the stimuli used in the experiment, categorized into three intact and three scrambled conditions: 1) unimodal videos of marmoset faces, 2) unimodal marmoset vocalizations, and 3) multimodal presentation of marmoset faces with corresponding vocalizations. Screenshots from the original recorded videos and custom-generated sound histograms, alongside custom-created icons for movies and sounds in PowerPoint, are by the authors. e Schematic of the sparse fMRI block design created by the authors, showing the temporal sequence of the presentation of intact and scrambled stimuli, each lasting 12 s, interspersed with 15 s baseline periods marked by a central fixation dot. During each run, the six conditions were presented in a randomized order and repeated four times, resulting in a total of 24 stimulus blocks and 25 baseline blocks. Each 3-second repetition time (TR) included a silent period of 1.5 s to ensure accurate perception of auditory stimuli.