Josefine A. Zerbe^1,8, Maggie Mae Mell^1,8, Juliane Damm², Marianna Schmidt², Nikolaus Weiskopf^2,3, Evgeniya Kirilina², Tomas Knapen^4,5,6,8, Martin N. Hebart^1,7,8; ¹Vision and Computational Cognition Group, Max Planck Institute of Human Cognitive and Brain Sciences, ²Department of Neurophysics, Max Planck Institute of Human Cognitive and Brain Sciences, ³Felix Bloch Institute for Solid State Physics, Leipzig University, ⁴Spinoza Centre for Neuroimaging, ⁵Netherlands Institute for Neuroscience, ⁶Experimental and Applied Psychology, Vrije Universiteit Amsterdam, ⁷Department of Medicine, Justus Liebig University Giessen, ⁸equal contribution

Previous research of human visual cortex has revealed functionally distinct and hierarchically organized processing pathways, each consisting of sequences of retinotopic maps. However, beyond early visual cortex the extent and even presence of these maps is not generally agreed upon. Specifically, it is unclear if variability in retinotopic structure reflects true idiosyncratic effects or noise in the retinotopic mapping data. Moreover, beyond the functional organization, the fine-grained structural connectome underlying human visual cortex is largely unknown and until recently had to be inferred indirectly from post-mortem studies. This has made it challenging to relate the macroscale structural organization of visual cortex to its functional topographic organization. To address these challenges, here we provide a densely-sampled dataset of 7 individuals, combining ultrahigh-resolution functional, structural, and diffusion data across 12 scanning sessions. We rigorously preselected individuals based on the reliability of an initial retinotopic mapping scan. Across 7 sessions of 7T MRI, we collected 192 minutes of retinotopy, as well as extensive object and motion-specific localizer and resting state data supplemented by multi-parameter mapping anatomy. Across three retinotopic mapping tasks, we (1) used moving bar apertures of varying width, (2) focused on the foveal representation (central 4 dva), and (3) focused on the periphery by moving the fixation cross to each corner of the screen. Across 5 additional sessions of 3T Connectom MRI, we further collected structural and diffusion data at 0.8mm isotropic resolution, allowing for a detailed mapping of short association fibers between adjacent brain regions. Data quality analyses of the functional data revealed minimal head motion and high noise ceilings, offering detailed, individually specific retinotopic maps. Paired with the densely sampled ultrahigh-resolution diffusion data, this dataset promises a highly detailed understanding of the functional neuroanatomy of the human visual system.

Acknowledgements: This work was supported by a Max Planck Research Group Grant and the ERC Starting Grant COREDIM awarded to MNH.