Abstract
Introduction: We used fMRI adaptation to investigate cortical selectivity to S-cone isolating (BY) and achromatic (Ach) stimuli. Previous work (Mullen et al, EJN, 2015, doi: 10.1111/ejn.13090) shows that for red-green (RG) and Ach contrast, there is increasing RG color selectivity in the higher ventral areas, especially VO. Here we used a similar paradigm to test the selective of responses to BY/Ach stimuli. Methods: We measured BOLD adaptation (3T scanner, TR=3s, 1.5 or 3mm isovoxels) to BY/Ach stimulus pairs (n=12), using similar methods to Mullen et al 2015. Both adapting and test stimuli were sinewave counter-phasing rings (0.5cpd, 2Hz), presented in a counterbalanced block design of adapt/no adapt, test and fixation blocks, with 96 repeats/subject. We used standard retinotopic mapping and localisers to define 9 ROIs (V1, V2, V3, V3a/b, LO, hMT, hV4, VO1 and VO2) and analysed data using AFNI/SUMA. Results: Across visual cortex, we found robust adaptation for all adaptor/test stimulus combinations (BY/BY, Ach/Ach, BY/Ach and Ach/BY). Across the 9 ROIs there was no significant main effect of either test or adaptor stimulus, but there was a significant interaction between these effects (F(1,385) = 9.23, p< 0.01). Ach test stimuli have a greater signal loss following BY than Ach adaptation, while BY tests had similar signal loss for both adaptors. Interestingly, the interaction was in the opposite in direction to that expected, with greater cross-stimulus adaptation than within-stimulus adaptation. Conclusion: Our data suggest that the S-cone pathway has qualitatively different interactions with cortical responses to Ach contrast than to RG. Responses to BY and Ach contrast are unselective across the cortical areas tested, unlike previous results for RG/Ach stimuli. The cross-adaptation of S-cone isolating stimuli on achromatic responses reveals an unexpected non-linear effect that is not accounted for by conventional stimulus adaptation models.
Meeting abstract presented at VSS 2018