Fig. 4.

Results of mixed-effects GLM analyses on the down-sampled oscillatory envelopes at each voxel. Maps of corrected z-statistics are shown in radiological view (axial, coronal and sagittal views from left to right). Z-statistics were corrected using false discovery rate (FDR) methods in FSL and subsequent z-statistical maps thresholded between 3 and 5 for visualisation, except the task blocks only 4–8 Hz (2 < z < 4) and 8–20 Hz (4 < z < 5.5) maps. An FDR of 0.05 was used for all the GLMs except the 4–8 Hz task blocks only analysis (in the white dashed box) where a FDR of 0.2 was used as no activity survived corrections at a FDR of 0.05. (a). Result of combined task and rest GLMs testing for differences between the 30 s task and rest blocks for the 4–8 Hz, 8–20 Hz and 20–40 Hz bands. (b) Result of task blocks only GLM testing for changes between the trials where a target was presented against all other non-target trials. In all GLMs, the average eye-blink time course was included as a nuisance regressor and hence all statistical inferences are orthogonal to eye-blink activity.