Ocular Dominance Columns Examined using BOLD fMRI with Phase Regression at 7 T

We investigated the signal spatial specificity of the GRE-BOLD contrast with and without phase regression in ocular dominance columns and layers of human V1. Phase regression is a post-processing technique that uses phase data to suppress the large vein signal, thus improving the GRE signal spatial specificity. Our results showed that the phase regressed laminar BOLD signal profile peaks towards the middle cortical depth, while that of GRE-BOLD was biased towards the cortical surface. Phase regression did not improve the contrast between columns, suggesting that the phase is not sensitive to intracortical veins running parallel to the cortical surface.