Fernanda L. Ribeiro¹ (), Torin Bambridge-Lozan¹, Noah C. Benson², D. Samuel Schwarzkopf³, Alexander M. Puckett¹, Steffen Bollmann¹; ¹University of Queensland, ²University of Washington, ³University of Auckland

The visual system comprises several functionally specialized cortical visual areas, where adjacent neurons represent adjacent retinal locations. These retinotopic maps are typically defined in polar coordinates, resulting in two orthogonal coordinate maps: one representing polar angle and the other eccentricity. While the retinotopic organization of early visual areas (V1, V2, and V3) in the human visual cortex is generally assumed to be organized according to a universal topological template that is similar across people, recent investigations have revealed compelling evidence of interindividual topological differences (Ribeiro et al., 2023; DOI:10.7554/eLife.86439). These differences cast doubt on the traditional template of early visual cortex organization. Therefore, we propose a unified, automated solution for retinotopic mapping and visual cortex parcellation based only on anatomical data derived from a T1-weighted image and that is not dependent on any single template of retinotopic organization. Our toolkit integrates (1) standard neuroimaging software (FreeSurfer 7.3.2 and Connectome Workbench 1.5.0) for anatomical MRI data preprocessing, (2) a deep-learning model (Ribeiro et al., 2021; DOI:10.1016/j.neuroimage.2021.118624) for predicting retinotopic maps at the individual level, and (3) an efficient implementation of the visual field sign analysis (Sereno et al., 1994; DOI:10.1093/cercor/4.6.601) for early visual areas parcellation. These components are packaged into Docker and Singularity software containers, which can be easily downloaded for local use and are available on Neurodesk (Renton et al., 2023; DOI:10.1101/2022.12.23.521691). Our toolkit can generate detailed, individual-specific retinotopic maps. Moreover, with polar angle and eccentricity maps, our toolkit generates visual field sign representations with unambiguous boundaries between early visual areas. These results demonstrate the potential of our open-source toolbox (https://github.com/felenitaribeiro/deepRetinotopy_TheToolbox) for individual-specific visual cortex parcellation.

Acknowledgements: This work was supported by the Australian Research Council (LP200301393) and the UQ AI Collaboratory.