Sergiu,
Your method would also work fine, if you also combine the bvecs files to
reflect the combined data file. Only downside would be that you might use a
lot more RAM for some preprocessing steps (e.g. DTIFIT) as more data would
need to be loaded into memory. But if you have the memory to spare and
differing bvecs, your method would be best.
Peace,
Matt.
-----Original Message-----
From: FSL - FMRIB's Software Library [mailto:[log in to unmask]] On Behalf
Of S. Groppa
Sent: Monday, January 21, 2008 7:53 AM
To: [log in to unmask]
Subject: [FSL] Process all or average
Dear list, dear Experts,
I was wondering if my processing pipeline is correct...and which is
the difference to that, discussed recently
I've got 3 averages (1 b=0 + 32 directions) so I do:
- fslmerge Input1 + Input2 + Input2 to a 4D_file
- eddy on this 4D_file 0
- DTIfit on this 4D file -
The histogram, FA, V1 and the visual) analysis are not bad but... Do I miss
smth??
I did the averaging as described in the list, but the results were not
good - besides differ my bvecs slightly (Inp1, 2,3 )...
Will be very grateful for any help
with best wishes
Sergiu
|