Hi again,
Thanks Johannes for the information about changing the file format in FSL. I
did that and did indeed get a stripped image, using BET, in dual file format I needed for input into SPM2. The stripped file can be read using MRIcro, however when I try to view the image using the SPM viewer, I cannot see any brain voxels in the viewer and the origin has become 2.3e+004,3.23e+004, 1.84e+004. And, I get this warning in the Matlab window:
SPM2: spm_image (v2.20) 14:09:43 - 02/11/2005
=======================================================================
SPM2: spm_image (v2.20) 14:10:00 - 02/11/2005
=======================================================================
Warning: Assuming a scalefactor of 1 for "C:\colin_brain.img".
> In spm_vol_ana at 129
In spm_vol>subfunc at 93
In spm_vol>subfunc1 at 62
In spm_vol>subfunc2 at 51
In spm_vol at 37
In spm_image at 302
In spm_image at 65
Any suggestions for how to resolve this issue? Thanks in advance,
Veronica
On Wed, 2 Nov 2005, Johannes Klein wrote:
RE: using nifti-1 with SPM2
> Hi Veronica,
> FSL determines the file type to write using the FSLOUTPUTTYPE variable. If
> you set this to ANALYZE, there is no need for an additional conversion step
> after stripping. Unfortunately, the orientation information in your .hdr
> file my change, but you can get around that by simply overwriting the new
> header with the old one:
>
> export FSLOUTPUTTYPE=ANALYZE
> bet original stripped
>
> and if you can't read the stripped in SPM:
> cp original.hdr stripped.hdr
> which overwrites your new header with the old one.
> Thus, you should end up with something you can use directly. BTW, if you use
> csh, you need to type "setenv FSLOUTPUTTYPE ANALYZE" instead of the "export"
> command. Please do check if the stripped image is L/R correct in SPM before
> further processing.
>
> Johannes
>
>
>
> Veronica Smith schrieb:
>> Hello,
>>
>> I am unable to find adequate information regarding using nifti file format
>> in SPM2. I used FSL 3.2b to strip the skull from colin27 brain image and
>> want to use
>> that for input into SPM2 for segmentation and normalization. FSL outputs a
>> .nii file which I converted to ANALYZE two-file format using MRIcro. When
>> I
>> input the stripped colin image(target) and the MNI152 stripped
>> image(source), I get an error "There is not enough overlap in the images
>> to
>> obtain a solution." When I view the colin image, there is nothing in the
>> viewer. Is there a way to resolve this file format problem? Perhaps I need
>> to use an
>> older version of FSL? Or is there a better way to convert the *.nii to
>> *.hdr/*.img format that SPM2 will read correctly?
>>
>> Thanks in advance for your help with this issue.
>> Veronica
>>
>
>
|