 |
 |
Hi, The inputs to randomise must specify images that exist in your filesystem and have correct paths so that they can be found from the current directory. It sounds like you are not using the correct names for the input files to randomise. For randomise there is no convention that you have to follow, so you can use any names that you want, as long as they represent valid image files. If you are unfamiliar with UNIX we recommend that you go through the first five sections of this tutorial: http://fsl.fmrib.ox.ac.uk/fslcourse/unix_intro/ All the best, Mark On 1 Aug 2013, at 20:58, Swetha Uppalapati <[log in to unmask]<mailto:[log in to unmask]>> wrote: Hi, I'm trying to perform statistical analysis of FA images using randomise. I have four different subjects with FA images that were created using this website: http://wikis.la.utexas.edu/imagelab/book/how-prepare-your-dti-images-fractional-anisotropy-fa-analysis Then I used http://wikis.la.utexas.edu/imagelab/book/statistical-analysis-fa-values to do the analysis. It says that you need the FA skeletonised image in the stats directory, but when I'm comparing 4 subjects how should I rename the FA images so that I can compare all of them. I have to rename them or else they would overwrite each other because they have the same name under each of the subjects stats directory. When I tried this command: randomise -i all_FA_skeletonised -o tbss -m mean_FA_skeleton_mask -d design.mat -t design.con -n 500 --T2 -V an error occurred saying FA image volume couldn't be read when I had a number or letter in front of the "all_FA_skeletonised" and "mean_FA_skeleton_mask" images in order to distinguish between the subjects. I also tried putting a * in the command but it had the same error. It could only read the image when it was the original name of the image with no letter or number in front. How should I save the images so that this problem doesn't occur? -Swetha