Dear FSL Experts:
I am analyzing data with randomise (randomise -i $input -o ROIs/$ROI/$ROI'_result' -d glm/glm_controls_patients.mat -t
glm/glm_controls_patients.con -V -x -T -m /home/shirer/avg152T1_gray_bin.nii) and am unsure how to interpret the corrected results. My uncorrected maps show robust results (in the areas where we expected to find group differences: p < 0.0001, extent > 1200 voxels), but the tfce-corrected and corrp maps are completely blank (range of 0 to 0). I was under the impression that randomise outputs pvalues for each voxel in the whole-brain maps (i.e., vox_p and tfce_p), and the maps are manually thresholded to determine whether anything reaches significance. Does randomise output blank t-maps if nothing survives the correction, or might there be something incorrect about the data input, contrast, or how I am executing the analysis?
The contrast is a simple two-sample t-test:
1 1 0
1 1 0
1 1 0
1 1 0
1 1 0
1 1 0
1 1 0
1 1 0
1 1 0
1 1 0
1 1 0
1 1 0
1 1 0
1 1 0
1 1 0
1 1 0
1 1 0
1 1 0
1 1 0
2 0 1
2 0 1
2 0 1
2 0 1
2 0 1
2 0 1
2 0 1
2 0 1
2 0 1
2 0 1
2 0 1
2 0 1
2 0 1
2 0 1
2 0 1
2 0 1
2 0 1
2 0 1
2 0 1
2 0 1
2 0 1
C1: 1 -1
C2: -1 1
Thank you very much for your help and guidance.
-Will
|