Hi,
I'm not sure - maybe you don't have good correspondence between the
masking of the two pre-aligned images. I would recommend trying FAST4
though as we are moving over to that in general.
Cheers.
On 18 Jun 2008, at 18:59, Hege Erdal wrote:
> Hi,
> I'm new with fsl, and have just installed fsl on my Ubuntu.
> I'm trying to use mfast, have used Bet to extract out non-tissue,
> but in a
> most cases, I get the error-message. I'm using mfast on a t1 and a
> t2 picture.
>
> /usr/share/fsl/bin/mfast -s 2 -c 3 -od
> /media/LaCie/NIFTI-format/Dykkar_Bo_nei/Dykkar_Bo_brain
> /media/LaCie/NIFTI-format/Dykkar_Bo_nei/Dykkar_Bo_brain
> /media/LaCie/NIFTI-format/Dykkar_Bo_nei/Dykkar_Bo2_brain.nii.gz
> No convergence in 30 SVDCMP iterations
>
> MFAST - Multispectral FAST Version 3.53
>
> Channel 0: /tmp/fsl_ASbhq4_Dykkar_Bo
> Channel 1: /tmp/fsl_ASbhq4_Dykkar_Bo2
> Imagesize : 256 x 256 x 19
> Pixelsize : 0.898438 x 0.898438 x 6.65
>
> Slice is too thick, 2D segmentation will be performed!
> initial segmentation by KMeans....
> 4 main iterations ...
>
> Matrix Inverse: Not invertable!
>
> Finished
>
> In the rest of the cases I get
>
> /usr/share/fsl/bin/mfast -s 2 -c 3 -od
> /media/LaCie/NIFTI-format/Dykkar_Df/Dykkar_Df_brain
> /media/LaCie/NIFTI-format/Dykkar_Df/Dykkar_Df_brain
> /media/LaCie/NIFTI-format/Dykkar_Df/Dykkar_Df2_brain.nii.gz
> 313577.5 681402.9 485882.3 1480862.8
>
>
> MFAST - Multispectral FAST Version 3.53
>
> Channel 0: /tmp/fsl_5nFAIf_Dykkar_Df_brain
> Channel 1: /tmp/fsl_5nFAIf_Dykkar_Df2_brain
> Imagesize : 256 x 256 x 19
> Pixelsize : 0.898438 x 0.898438 x 6.65
>
> Slice is too thick, 2D segmentation will be performed!
> initial segmentation by KMeans....
> 4 main iterations ...
>
> Segmentation done successfully!
>
> Calculation time 51 seconds!
>
> Write segmentation image /tmp/fsl_5nFAIf.fastout/QQQQQ_seg.hdr
> Successfully!
>
> Class: tissue 0 tissue 1 tissue 2 total
> Volume:
> Finished
>
>
> Do you have any idea what I can do to fix this? I ran the fsl
> evaluation,
> and it said that the tests passed. I think maybe there might be
> something
> wrong with the brain extraction, cause when I tried to use an another
> computer with mfast, it all worked out fine.
>
> Thanks,
> Hege
>
---------------------------------------------------------------------------
Stephen M. Smith, Professor of Biomedical Engineering
Associate Director, Oxford University FMRIB Centre
FMRIB, JR Hospital, Headington, Oxford OX3 9DU, UK
+44 (0) 1865 222726 (fax 222717)
[log in to unmask] http://www.fmrib.ox.ac.uk/~steve
---------------------------------------------------------------------------
|