Hi,
in addition to eddy current correction, watch out that when using analyze pairs your images are uniformly scaled (SPM used to introduce a volume-specific scaling).
Cheers-
Andreas
________________________________
Von: FSL - FMRIB's Software Library im Auftrag von Ed Roberts
Gesendet: Di 27.06.2006 17:59
An: [log in to unmask]
Betreff: [FSL] FA pipeline
Hi,
I'm trying to generate some FA maps to do a regression analysis with, and I
wanted to check whether I was using the right things in the right order.
I'm using a Siemens 1.5T scanner with 12 directions, b-values 0 & 1000.
1. Dicoms from the scanner into SPM to get hdr and img files.
2. All 13 diffusion files into MRICRO to get a 4D data file.
3. Normalise B-vec values.
4. B0 image into BET to get a no_diff and no_diff_mask file.
5. Run dtifi.
So, I'm concerned that I'm missing out doing any registration or alignment,
or does this happen in FDT? Apparently the data I need to adjust for the
eddy currents isn't easily available, so I might have to skip that.
Any my other slightly related question is how do I know if Bedpost has
crashed? Is there any good way to tell or do you just have to cross your
fingers and wait?
Thanks,
Ed Roberts
|