Dear FSLers:
We would like to implement TBSS in our FA analysis, and have some questions
as we can still adjust our DTI protocol to our needs. I’m still
inexperienced in this field so I apologize if these questions are
elementary, or have been answered repeatedly. Also, please correct me if
I’m wrong.
To use TBSS,
Q1: FA is the only modality required by TBSS.
Q2: DWIs do not have to be acquired more than 25 directions like
tractography, and the voxel dimensions can be isotropic(2x2x2mm^3 and
4x4x4mm^3 Page7 in technical note), or unisotropic like what we have
(1.6x1.6x4mm^3 and 1x1x5mm^3)?.
I’m hung up on FAQ in FDT. If isotorpic voxel size is desirable, can we
reslice FA? I have no background to justify this though.
Q3: In pg12 of technical note results section, I only see isotropic voxel(I
assume no interpolation from acquisition) taken at ~60 directions! and
multiple repeats. We were told that we can only acquire DWIs along 12
directions at this point with our Siemens scanner. Are 12 directions
sufficient to acquire decent FA and run TBSS? We can acquire b0 multiple
times so I can follow the guideline.
Q4: Is it correct by saying that multiple b0s and repeats average out
small fluctuation in signals so that it enhances overall signal to noise
ratio of FA?
PC spec:
3.0Ghz(x2) Xenon with 6G RAM running RedHat 32-bit WS 4.0
We acquire DTIs with two scanners
This protocol is adjustable, so please advise us if there is any suggestion.
Siemens 3.0T Allegra
b=1250
12 directions of DWIs+1 b0
TR/TE=5100msec/8msec
Reconstructed voxel dimensions: 1.64x1.64x4mm(128x128x32 slices)
Not sure about the slice gap.
GE 1.5T Signa
b=1000
25 directions of DWIs +1 b0
TR/TE=8300msec/100.8msec
Reconstructed voxel dimensions: 1.02x1.02x5mm (256x256x26 slices).
Unfortunately this one is interpolated, and there is nothing we can do.
Not sure about the slice gap.
Thanks in advance
Hedok
|