Marianne Duyck¹ (), Kurt Braunlich², Chris, I. Baker², Bevil, R. Conway¹; ¹National Eye Institute, ²National Institute of Mental Health

The macaque monkey is commonly used as a model for understanding human visual processing. However, methods for comparing the two brains have been limited, and we still know little about the changes in processing and organization that have occurred over the 25 million years of evolution between them. Here, we focus on retinotopically organized cortex. Even for early visual areas, the homology between species is disputed, such as the apparent absence of a separate dorsal V4 in humans. To address these gaps, we used a common space model approach (also known as hyperalignment) on fMRI data collected on 3 monkeys and 9 humans viewing the same videos. Using an atlas-based definition of early visual areas, we learned and tested a cross-species model across pairs of early visual areas from both hemispheres. Our results indicate significant between-species shared information in all pairs of areas, which is higher for pairs from the same hemisphere, higher for areas with the same label (V1 with V1, V2 with V2, etc.), and decreases from posterior to anterior: with the highest cross-species similarity for V1, and the lowest for MT. Additionally, we collected data to delineate retinotopically organized areas in the same subjects using population receptive field mapping (pRF). Ongoing work is using these functionally defined regions as seed regions for the common space model. This approach places us in a unique position to address ongoing controversies regarding the homology and organization of visual areas across the two species.