 |
 |
Hi Everybody, I'm so sorry to beat a dead horse, but I've been reading through the archives for days regarding the best way to extract average FA values for specific ROIs and I'm still unsure on the process. I'd like to use the JHU-ICBM WM labels atlas for the 48 predefined ROIs and then I'd like to extract average FA values from my all_FA_skeletonised.nii.gz file (in order to extract each individual's average FA for a given ROI). I'm a bit confused on the best way to do this. From my understanding, both the JHU-ICBM WM atlas and the all_FA_skeletonised are in MNI space, correct? If that's true, does this process seem to be anywhere close to correct? After reading some of the posts, is this how I would create a mask for each separate region (eg. splenium)? 1. fslmaths /usr/local/fsl/data/atlases/JHU/JHU-ICBM-labels-1mm.nii.gz -thr 5 -uthr 5 -bin splenium.nii.gz * this command just selects the ROI, right(ie. 5=splenium)? 2. fslmaths splenium.nii.gz -mas mean_FA_skeleton_mask.nii.gz -bin spleniummask.nii.gz * does this make it so that the mask only takes voxels that are included in my mean_FA_skeleton_mask as well (ie. the intersection of the splenium mask from the previous command and my entire sample's mean FA skeleton mask)? 3. fslmeants -i all_FA_skeletonised.nii.gz -m spleniummask.nii.gz * I assume this is just applying that splenium mask to every volume (subject) of the all_FA_skeletonised file? Also, does the fslmeants command just give me the mean of ALL subjects as opposed to the mean FA of each subject within the specified mask? 4. fslstats all_FA_skeletonised -k spleniummask.nii.gz -M -t > splenium.txt * I want to make it output the mean FA within the defined ROI of each subject into a txt file, is this even close? Out of curiosity, is there a better or more efficient way to do this? I see that there are maxprob tract files as well...would I be better off creating the masks from the JHU-ICBM-maxprob-thr25-1mm instead (or a different probability threshold)? Just so you know, I'd like to eventually make a large file with the average FA of each individual for all of the JHU-ICBM 48 ROIs so that I can analyze things within whatever statistical package we choose. Thank you so much for your help and sorry for the lengthy post. Best Regards, Jon