Ekin Tünçok¹ (), Lynne Kiorpes¹, Marisa Carrasco¹; ¹New York University

[Background and goal]. Visual discriminability varies with polar angle at isoeccentric locations. In human adults, performance is typically higher for stimuli along the horizontal than the vertical meridian (horizontal-vertical anisotropy, HVA), and along the lower than the upper vertical meridian (vertical meridian asymmetry, VMA). The macaque visual system is a prevalent animal model for understanding human vision. However, the natural-scene statistics may differ between these two species due to differences in bipedal locomotion and interactions with peri- and extra- personal space. Here we investigate whether polar angle asymmetries differ between adult human and macaque observers. [Method]. Adult macaque (n = 6) and human observers (n = 20) completed a 2AFC motion direction discrimination task at isoeccentric locations (7°) around the visual field. On every trial, a vertically-oriented Gabor target was presented (500 ms) among 7 equidistant distractors (Gabor patches with varying motion directions). Observers reported the motion direction of the target (left or right) whose location randomly varied among the 8 isoeccentric locations. [Results]. Performance was not homogenous as a function of polar angle for either observer group. Interestingly, the two groups of observers showed a striking difference in sensitivity (d’): For human observers, there was a pronounced VMA, a weaker HVA, and lower sensitivity at intercardinal than cardinal locations. For macaque observers, sensitivity was more homogenous as a function of polar angle, but was the poorest at the lower vertical meridian, showing an inverted VMA. [Conclusion]. The macaque visual system is used as an animal model to understand human vision, yet these two observer groups showed a pronounced difference in sensitivity around the visual field. The inverted VMA observed in macaques may reflect adaptive behavior by increasing discriminability at locations with greater relevance for visuomotor integration.

Acknowledgements: Funding: NIH R01-EY027401 to M.C.