I am using qboot to create empirical distributions of bootstrap samples for FA. I am trying to figure out if there is a way to permute these distributions using a tool like randomise. We have bootstrap samples for the same subjects, pre and post- condition, and want to permute the original bootstrap samples of FA-pre and FA-post at each voxel location. The output would be a test statistic at each voxel. Any guidance on how to set up the permutations and if I can use randomise to do this?