Search Abstracts | Symposia | Slide Sessions | Poster Sessions
Advancing the precise and reproducible localization of cortical reading circuits using dense-sampled fMRI
Poster Session D, Thursday, October 1, 4:30 - 6:30 pm, Wangari Maathai
Garikoitz Lerma-Usabiaga1,2, Pedro M. Paz-Alonso1,2, Yongning Lei1; 1BCBL. Basque Center on Cognition, Brain and Language, San Sebastián, Gipuzkoa, Spain, 2IKERBASQUE. Basque Foundation for Science, Bilbao, Bizkaia, Spain
Reading is a culturally acquired cognitive skill that relies on a distributed cortical network, with the ventral occipitotemporal cortex (VOTC) serving as a critical hub for orthographic processing. While neuroimaging indicates multiple functionally distinct subregions within the VOTC dedicated to visual word recognition, their specific computational roles and exact anatomical borders remain subjects of intense debate. A major driver of this uncertainty is the lack of methodological standardization. Functional localizers vary drastically across laboratories in stimulus design (e.g., words vs. false fonts, scrambles, or objects), contrast selection, and region-of-interest (ROI) definition protocols. Consequently, this high variability yields inconsistent spatial coordinates in group-averaged templates and severely limits cross-study reproducibility. To resolve these methodological hurdles, we pursued a dual approach. First, we conducted an exhaustive synthesis of recent literature concerning category-selective ROI definitions within the VOTC. This review identified a consistent, yet previously unconsolidated, network of seven distinct reading-related subregions spanning the posterior-to-anterior occipitotemporal gradient. Second, to validate this model empirically, we executed a dense-sampling, high-resolution functional MRI (fMRI) study. We scanned 10 participants across 10 independent sessions each using multiple standard localizer paradigms. This approach yielded an exceptionally high signal-to-noise ratio (SNR) dataset per participant, allowing for highly precise, individual-level mapping of VOTC reading circuits without the spatial blurring inherent to group-average methods. Crucially, our empirical data confirmed the reliable, stable localization of all five specific subregions at the single-subject level. While macro-anatomical positions varied across individuals, each ROI exhibited a robust, highly idiosyncratic functional topology that remained remarkably stable across separate scanning sessions. Furthermore, leveraging this dense-sampled dataset, we built a robust power analysis framework designed to optimize future functional localizer paradigms. This framework systematically maps the mathematical trade-off between scan duration (number of execution runs) and ROI detection sensitivity, demonstrating how brief conventional localizers often fail to detect more anterior, effect-size-limited VOTC reading patches. Ultimately, this research offers a highly inter-subject reproducible map of VOTC reading circuits, establishes concrete experimental guidelines for fMRI data acquisition, and provides an optimized statistical framework for individual-level brain mapping. By offering actionable guidelines to maximize statistical power, this work elevates the methodological rigor of future reading research and provides a foundation for reconciling disparate findings across the neurobiology of language literature.
Topic Areas: Reading,