Search Abstracts | Symposia | Slide Sessions | Poster Sessions
Short, dynamic fMRI task localizes category-specific brain responses to language-related visual stimuli
Poster Session C, Thursday, October 1, 10:45 am - 12:45 pm, Wangari Maathai
Iliana I. Karipidis1,2,3, Chayenne Garcia1, Alexandra Schreiber1,2, Amelie Haugg1, Silvia Brem1,2,3,4; 1Department of Child and Adolescent Psychiatry and Psychotherapy, University Hospital of Psychiatry Zurich, University of Zurich, Switzerland, 2Neuroscience Center Zurich, University of Zurich and ETH Zurich, Switzerland, 3Competence Center of Language and Medicine, University of Zurich, Switzerland, 4University Research Priority Program (URPP), Adaptive Brain Circuits in Development and Learning (AdaBD), University of Zurich, Switzerland
The human visual system is functionally organized in category-selective regions that selectively respond to relevant categories, such as faces, bodies, places, and words (Rosenke et al. 2021). While some parts of this system in the ventral occipitotemporal cortex (vOTC) are assumed to have evolved to process visual categories that are evolutionary relevant, such as faces (Deen et al. 2017), other parts of the vOTC have been shown to be highly plastic and can develop to process visual categories that become relevant through experience and learning, such as written text (Dehaene et al. 2010). Several category-selective visual areas are relevant for language processing, including parts of the vOTC that selectively respond to faces, written text, and hand gestures. Here, we developed a short, dynamic fMRI localizer to identify regions selective for language-related visual categories. Healthy adults (N=37; age m=24.9±4.7; 31 females) completed a short, dynamic fMRI localizer task (5 min) including short (3s) visual clips of sentences (SE), falsefonts (FF), checkerboards (CH), gestures (GE), meaningless hand movements (HM), talking faces (FA), and scrambled image recordings (SC). Using a block design, each condition was presented in 3 blocks of 15s each. The fMRI data was preprocessed using fMRIPrep. Statistical analyses were conducted using Nilearn and SPM, with t‑contrasts computed to identify text-selective (SE vs CH; SE vs FF), gesture-selective (GE vs SC; GE vs HM), and face-selective responses (FA vs SC). Category-selective regions for sentences, gestures and faces were first localized in a mask of the vOTC on a single-subject level (p<0.05). In addition, whole-brain group analyses were performed (FWE corr p<0.05). We successfully localized distinct subregions of the left and/or right vOTC selective for sentences, gestures and faces at the individual level. Substantial interindividual variability was observed in the location, extent, and intensity of these responses. Group analyses revealed sentence-selective responses (SE>FF; SE>CH) within core reading network regions, including bilaterally the superior and middle temporal gyri (STG/MTG), the left inferior frontal gyrus (IFG), the left precentral gyrus, and lateral vOTC subregions known as the visual word form area. Gesture-selective responses (GE>SC; GE>HM), partially also overlapped with the reading network and were mainly found bilaterally in the IOG, MTG, vOTC, and superior parietal lobule (SPL). Face-selective responses (FA>SC) were found bilaterally in the amygdala, STG, MTG, angular gyrus and medial vOTC subregions, i.e. the fusiform face area. These results demonstrate that our novel, short dynamic localizer identifies language-related vOTC subregions at the individual level. Notably, sentence-selective vOTC subregions were often adjacent to gesture-selective vOTC subregions, consistent with prior findings for word and limb processing (Nordt et al. 2021). Group‑level results confirmed that the task engaged the neural circuits involved in processing text, gestures, and faces. This work illustrates the efficiency of dynamic functional localizers to generate strong responses in brain circuits of interest within short acquisition times (Arsalidou et al 2011; Küçük et al. 2024). This short dynamic localizer may present an engaging and child-friendly tool (Deen et al. 2017) to study the development of language-related visual processing in young children.
Topic Areas: Reading, Multisensory or Sensorimotor Integration