Hi Donovan,
If you want to overwrite the matrices in the .feat/reg subdirectory then
that's what will get used in any later analyses that need the transforms.
The important ones are the example_func2*.mat files. To find out how to
generate the gif images, if you want that, see the relevant commands
in report.com like
/usr/local/fsl/bin/flirt -ref standard -in example_func -out example_func2standard -applyxfm -init example_func2standard.mat -interp trilinear
and the two commands that follow this (convert_xfm and slicer).
Yes, the file report.com only tells you about external calls made by the
main TCL script, so anything (such as generating the web page) done by the
script itself won't get recorded there. But I think that stuff relevant to
registration should be in there. Not sure what you mean by binary garbage
- shouldn't be anything like that in there.
FSL doesn't use the (non-standard) co-ordinate origin to aid the
registrations, but does use it in the standard space image in order to get
"Talairach" co-ordinates for co-ordinate reporting.
Cheers.
On Fri, 5 Dec 2003, Donovan Rebbechi wrote:
> Hi. I'm having some difficulty with between-modality registration in Feat.
> With a 3 parameter model, the images are getting shifted substantially in
> the I have some questions related to this:
>
> (1) Is there a way to run the FEAT analysis stream in a way that bypasses
> inter-modality registration ? For example, in our data, we are able to get
> a good match by specifying an origin in BOLD and T1 images (what amounts
> to a "manual 3 paramater" registration). Is there are way to override the
> matrix that feat computes ? One thought I had was converting the
> report.com file into a shell-script (and editing the part that generates
> the transformation matrices for inter-modality reg), it looks like a
> record of what programs were run -- but it also seems to be cluttered
> with binary garbage, so I'm not at all clear if this file is an accurate
> record of what was run.
>
> (2) does FSL use the (non-standard) origin fields in the Analyze header ?
> I tried setting the origin fields (same as SPM uses) in hope that this
> would give the algorithm enough information to work with, but it still
> chokes.
>
> --
> Donovan Rebbechi, PhD
> Research Associate, Dept of Psychology
> Smith Hall #340, 101 Warren Street
> Newark NJ 07102
> 973 353 5440 x240
> http://pegasus.rutgers.edu/~elflord
>
Stephen M. Smith DPhil
Associate Director, FMRIB and Analysis Research Coordinator
Oxford University Centre for Functional MRI of the Brain
John Radcliffe Hospital, Headington, Oxford OX3 9DU, UK
+44 (0) 1865 222726 (fax 222717)
[log in to unmask] http://www.fmrib.ox.ac.uk/~steve
|