Pinglei Bao¹, Baoqi Gong¹; ¹Peking University

Visual information about food is crucial for survival in primates, yet the neural mechanisms for identifying and processing food stimuli remain unclear. Previous studies in humans have localized regions in the ventral temporal cortex that show a strong preference for food stimuli using fMRI datasets (Allen et al., 2021). However, the response of individual neurons within these regions has not been explored due to the lack of suitable animal models. To address this, we conducted a food localizer experiment with four macaques, comparing photorealistic and cutoff food images against corresponding non-food images. This analysis revealed a consistent food network across subjects, identifying three distinct food-selective regions spanning from posterior to anterior in the inferotemporal cortex. We then targeted the middle food region using Neuropixels probes for recordings. A high concentration of the food-preferred neurons was identified with two subtypes of food-preferred neurons: one responding more strongly to photorealistic food images and the other showing selectivity for cutoff food images. These neuron types also differed in cortical depth, with cutoff food-preferred neurons in superficial layers and photorealistic food-preferred neurons in deeper layers. To further investigate whether macaque food regions are homologous to human food areas, we calculated the correlation between mean neuronal responses and human fMRI voxel responses to the same set of images. Regions with high correlation significantly overlapped with human food areas. Overall, our results indicate that the macaque food area closely mirrors the human food network, suggesting it as a valuable model for investigating the neural mechanisms underlying food-related visual processing.