fMRI reveals a modulatory role of visual field meridians on object-based selective attention

Poster Presentation: Sunday, May 19, 2024, 8:30 am – 12:30 pm, Pavilion
Session: Attention: Features, objects 2

David H. Hughes1 (), Adam J. Barnas2, Adam S. Greenberg1; 1Joint Department of Biomedical Engineering, Medical College of Wisconsin and Marquette University, 2Department of Psychology, University of Florida

Previous behavioral results demonstrated a shift direction anisotropy (SDA) during object-based attentional selection that was modulated by visual field meridians (Barnas & Greenberg, 2016; 2019). We hypothesized that differential prioritization of horizontal and vertical target locations gave rise to the SDA due to attentional resources being divided differently by horizontal and vertical meridians. To test this, we measured BOLD signals during fMRI from visual cortex in 19 healthy adults. Subjects detected a target (letter T among distracting Ls) within an L-shaped object with two possible locations: vertex in upper-left (UL) or lower-right (LR) quadrant. Cue validity was 60% valid, 20 % invalid, 20% catch. Invalid trials were counterbalanced between shift orientation (horizontal and vertical) and whether target detection required a meridian crossing (crossing and non-crossing). The crossing SDA was significantly greater than the non-crossing SDA for the UL (p = .039, d = 0.60), but not the LR (p = .454, d = 0.04) object. A repeated measures ANOVA of cue-related activation revealed significant main effects of object location (p < .001) and retinotopic ROI (p < .001), as well as a significant interaction (p = .007). Post-hoc Tukey test for the UL object revealed no activation differences between valid and either non-crossing invalid-horizontal (p = .963, d = 0.17) or non-crossing invalid-vertical (p = .969, d = 0.15) locations, suggesting that attention pools within the cued quadrant. Compared to non-crossing, crossing activation was reduced (invalid-horizontal: p < .001, d = 0.95; invalid-vertical: p < .001, d = 0.59). This prioritization of each invalid location is consistent with the attentional spreading hypothesis (Richard et al., 2008). However, cue-related activation did not fully account for the SDA. Instead, the SDA could arise from differential spreading of attention resulting from stronger segregation of pools of resources by the vertical versus horizontal meridian.

Acknowledgements: This work was supported by grants from the US-Israel Binational Science Foundation (#2013400) and National Science Foundation (SBE 2122866).