LangCog Seminar - Xiangqi Luo

Object concepts can be accessed through different input modalities, such as pictures and written words. However, it remains unclear how conceptual categories interact with stimulus input modalities in the brain. By combining multimodal neuroimaging and electrophysiological approaches, including fMRI, sMRI, and SEEG, this study investigated the cross-modal representation of three object categories—animals, tools, and scenes—in the occipitotemporal cortex (OTC), a key region for category-specific object processing. First, task-based fMRI was used to identify category-specific regions in the OTC for pictures and written words, and to characterize their cross-modal representation profiles. The results revealed clear category differences: animal-specific regions were largely picture-specific, whereas tool- and scene-specific regions showed stronger cross-modal responses. Functional connectivity and coupling analyses further suggested that interactions between OTC category-specific regions and other brain areas may influence the degree of cross-modal representation, particularly for scene concepts. To further examine whether these tool- and scene-specific regions are necessary for cross-modal representation, sMRI-based lesion analyses mapped cortical damage in patients with brain lesions onto behavioural performance across picture and word tasks for different categories. These analyses revealed a causal contribution of the left parahippocampal gyrus to the cross-modal representation of scenes. Finally, SEEG recordings were used to characterize the temporal dynamics of category-specific representations across different modalities. The results showed that although some regions exhibited cross-modal responses, the timing at which these representations emerged varied across modalities. In summary, this study provides a multidimensional account spanning cross-modal representation profiles, influencing factors, neural necessity, and spatiotemporal dynamics, thereby deepening our understanding of how conceptual categories interact with stimulus input modalities in the brain.

Dr Xiangqi Luo received her PhD from Beijing Normal University in 2025 and is currently a postdoctoral research fellow at the LRNLab in UCL PALS. Her research focuses on the structural and functional organization of the ventral occipitotemporal cortex (VOTC) and how this region supports visual object and word processing.