Zelin (Linda) Zhao¹ (), Noah C. Benson², Jan W. Kurzawski¹, Denis Pelli¹, Jonathan Winawer¹; ¹New York University, New York, NY, USA, ²University of Washington, Seattle, WA, USA

Crowding is the failure to identify a visual target due to nearby clutter. It is a major bottleneck for recognition, affecting reading and visual search. Despite its importance both clinically and practically, the neural basis of crowding is not well established. Crowding distance is the minimum separation between a target and surrounding objects that allows recognition. It increases with eccentricity. This has been approximated as a linear function (Bouma law). Our recent study, however, found that the increase is nonlinear, captured by a polynomial with a quadratic term (Kurzawski et al, 2023, Journal of Vision). Here, we take advantage of this finding to ask whether the increase in crowding distance with eccentricity is related to cortical magnification of retinotopic maps. More specifically, we hypothesized that crowding distance, when projected to mm spacing on a cortical map, is a constant. We test this conservation hypothesis separately in 4 maps: V1, V2, V3 and hV4. To do so, we measured crowding distance at 2.5, 5, and 10 deg eccentricity using psychophysics, and measured retinotopic maps with functional MRI. We then calculated cortical crowding distance, measured in mm of cortical spacing, as the product of crowding distance, in deg, and cortical magnification, in mm per deg. We find that in hV4, cortical crowding distance is conserved with eccentricity: 1.5±0.08 mm, 1.6±0.15 mm, and 1.8±0.15 mm for 2.5, 5, and 10 deg targets. In contrast, in V1 to V3, the cortical crowding distance increases systematically with eccentricity. Our finding that cortical crowding distance is conserved across eccentricities in the hV4 map complements our recent finding that it is also conserved across observers in the hV4 map. Together, the results show that crowding is closely linked to the cortical magnification of the hV4 map.