Robert Satzger^1,2,3, Tomas Knapen^4,5,6, Martin N. Hebart^1,2,3; ¹Department of Medicine, Justus-Liebig University Giessen, Giessen, Germany, ²Vision and Computational Cognition Group, Max Planck Institute for Human Cognitive and Brain Sciences, Leipzig, Germany, ³Center for Mind, Brain and Behavior (CMBB), University of Marburg, Giessen and Darmstadt, Germany, ⁴Spinoza Centre for Neuroimaging, Amsterdam, The Netherlands, ⁵Department of Computational Cognitive Neuroscience and Neuroimaging, Netherlands Institute for Neuroscience, Amsterdam, The Netherlands, ⁶Department of Experimental and Applied Psychology, Vrije Universiteit Amsterdam, Amsterdam, The Netherlands

Retinotopic mapping is essential for determining the cortical topography of visual areas and delineating their boundaries. Despite its general usefulness, retinotopic mapping often suffers from two challenges. First, retinotopic mapping is typically time-intensive, but scan time is limited and expensive. Consequently, researchers would benefit from a systematic analysis of how much scan time is sufficient for stable retinotopic maps. Second, head-coil and projector geometry often place hard constraints on the field of vision inside the scanner, leaving most of peripheral visual cortex unmapped. Here, we address complementary solutions to both problems. For challenge 1, we quantify how map quality scales with scanning time to determine when performance saturates. For challenge 2, we introduce a corner-fixation paradigm that effectively expands the mappable visual field: by shifting gaze toward each corner of the screen, the stimulus reaches regions otherwise outside the displayable area. Combined, this almost quadruples the mappable visual field. We analyzed two 7T fMRI datasets, one consisting of 10 participants from the HCP retinotopy dataset (Benson et al., 2018), the other consisting of seven participants from the denseRetinotopy dataset (Zerbe et al., 2024). For challenge 1, our results surprisingly showed that a single run of ~3 min was sufficient to produce qualitatively clear polar angle and eccentricity maps at 7T that already stabilized in early visual areas, with only minimal further improvement for longer acquisitions. For challenge 2, we found that corner fixation expanded the mappable eccentricity from 10° to 18° vertically and 17° to 32° horizontally, while yielding qualitatively comparable maps to centrally presented stimuli. Together, these findings demonstrate that it is possible to strongly improve the efficiency of retinotopic mapping at ultrahigh field strengths by trading shorter acquisition duration with increased visual field coverage.