I would like to calculate the respective (normalised for ICV) volumes of the caudate and putamen in standard space (MNI152 1mm) using fslstats. I have used FIRST and the run_first_all script (with -b) to segment subcortical structures. I then used FLIRT and applied the transformation *_mr_to_std_sub.mat to *mr_all_fast_firstseg.nii.gz:
flirt -in "$INFILE" -ref "$BASE"_mr_to_std_sub.nii.gz -out "$BASE"_mr_all_fast_firstseg_to_std_sub -init "$BASE"_mr_to_std_sub.mat -applyxfm
However, the respective masks for each structure in the output file "$BASE"_mr_all_fast_firstseg_to_std_sub.nii.gz now contains a "grainy" outline with dissolved borders between structures so that I cannot use thresholding function in fslstats to calculate volume/#voxels.
1. How can I transform *mr_all_fast_firstseg.nii.gz to standard space and retain the integrity of the mask of each structure?
2. Would it be appropriate to use the *structure_corr.nii.gz and just apply the *_mr_to_std_sub.mat in FLIRT?