Distributed representational encoding of food attributes in ventral visual cortex

Abstract

Despite substantial progress in understanding how visual features of food are processed in the brain, it remains unclear how subjective food properties, such as perceived palatability, perceived caloric content, and perceived health value, are reflected in neural representational structure after accounting for visual similarity. Using functional MRI and representational similarity analysis (RSA), we examined visual, subjective, and categorical food representations in 25 healthy young women viewing 96 food images during an orthogonal color-discrimination task. Univariate analyses revealed reliable activation differences between high- and low-calorie foods in visual and ventral temporal cortex, with additional effects in orbitofrontal regions. Surface-based parcel-wise RSA and hypothesis-driven ROI analyses showed that visual models explained representational structure most strongly in early visual cortex and extended into occipitotemporal regions. Food-related models showed smaller but reliable effects in lateral and ventral occipitotemporal cortex: perceived calorie, perceived health, and objective calorie category were all associated with neural representational geometry in LOTC and VOTC. However, partial RSA and commonality analyses indicated that these effects were largely shared rather than separable. Perceived-calorie structure remained reliable after controlling for visual models, but not after controlling for perceived health. Palatability did not show reliable positive representational correspondence. These findings suggest that food-related information is reflected within occipitotemporal visual representational spaces during incidental viewing, but not as an isolated scalar calorie code. Instead, calorie-related effects appear to reflect a broader food-property axis on which energy density, perceived health, category structure, and learned food knowledge covary.