Apply TBSS FA analysis results to non-FA images

The TBSS analysis on the FA data the warping information can be applied to the RD, AD, and MD diffusion metrics. Like the TBSS FA processes, the TBSS scripts for the other metrics must each be run from the directory in which the images are stored, and the non-FA files must have the exact same file name as the corresponding FA file for each subject. As such we created a subdirectory for each of the three non-FA metrics (AD, MD, and RD), and copied the relevant image file (e.g. AD images went into the AD TBSS sub-directory) but changing the filename to the subjects' corresponding FA file name, e.g. the file called origdata/Subject001_AD.nii.gz was copied and changed to AD/Subject001_FA.nii.gz. Note that the RD image files were calculated as a mean of the two eigenvectors that are the non-principle diffusion direction.
With the dti image files in their respective subdirectories and named the same as the corresponding FA file for each subject, the TBSS nonlinear registration was applied, merging all subjects' warped data into 4D image files--stats/all_AD, stat/all_MD, stats/all_RD-- and projecting this onto the original mean FA skeleton, resulting in the 4D projected data--stats/mean_AD_skeletonised, stat/mean_MD_skeletonised, and stat/mean_RD_skeletonised. The code uses the "tbss_non_FA" script and telling what the alternate data is called (e.g. AD, MD, or RD).