Hi all,
I need to normalise a timecourse of images to an ROI/mask covering a
specific piece of anatomy. So what I want is similar to:
fslmaths input -inm 10000 output
except that rather than the mean of the whole image set to 10000, I want the
mean of my ROI set to 10000 (with all other voxels in the image at that
timepoint being scaled to maintain the relative SIs). So across the entire
timecourse the mean of the ROI would be fixed to 10000.
The only way I can think of to do this is something along these lines:
1)fslmeants to output a text file of SIs describing that ROI over time
2)fslsplit to create 3D vols
3)script fslmaths to do this: fslmaths input3D_Tx -mul (10000/(value x in
fslmeants.txt)) output3D_Tx
4)fslmerge to put the data back together again.
Is there a simpler way I have missed?
Thanks for your help,
Nad
|