Dear Guillaume,
Thank you very much for the prompt answer and the script!
I have just few questions about it:
1. I am not sure why you indicated two mask images (mask.nii and mask_from_atlas.nii) and not only the inferior temporal mask (that I suppose is the mask_from_atlas.nii), why the spmT_0001.nii needs to be filtered by both masks?
2. You suggested to use the spmT_0001.nii image as SPM Results image instead of the con_0001.nii image and the image I proposed in the question [Results → SPM.mat → select contrast → unc p<0.005 → save... all clusters (n-ary)], could you explain it a bit why?
For what I know the spmT image contains a t-statistic value at each voxel. When I select the 500 voxels with the highest value, there might be voxels with a t value below the T threshold (in case there are not 500 voxels above the T threshold for a particular subj/contrast). Is there a way to prevent this? For this reason, I thought to use the contrast image resulted by saving all clusters above the threshold I defined in the Results (in my case unc p<0.005), but maybe it is not the correct procedure.
How can I check if the 500 most active voxels within the mask have a t-statistic value above the T threshold (calculated from a unc p<0.005)?
Best wishes,
Alessio
|