It's possible that this one subject was converted on a different
platform and has the byte ordering wrong compared with the other
images. Another possibility is that areg/nreg crashed due to
mathmatical errors; this happens very rarely.
Cheers, Steve.
On 29 Jan 2008, at 20:15, Dianne Patterson wrote:
> Dear Group,
> Aha...I found the proximate cause of my problem.
> The *hr.nii.gz image for one subject (one in the middle, not at the
> beginning or end of the processing) was empty.
> When I moved his files out of the FAi directory and reran
> tbss_3_postreg, I get all_FA and the skeleton image and they look
> fine.
>
> Now, that one subject looks fine in the *.png in slicesdir...(the
> brain is right way around and looks to be sliced in roughly the same
> places as everyone else's brains.
> His other *.nii.gz image looks okay, and the original img and hdr
> look okay...
>
> Is there some way for me to determine why this fellow's *hr.nii.gz
> image was empty?
>
> Thankyou,
>
> Dianne
>
>
> On Jan 29, 2008 12:31 PM, Dianne Patterson <[log in to unmask]>
> wrote:
> Thanks Gwenaelle for your suggestions.
>
> I removed the stats directory and reran tbss_3_postreg.
> I get 14 empty volumes in all_FA.nii.gz...and, as you can see, no
> obvious error messages.
> I'm going to go dig around in the scripts a little and the existing
> images, to see if I can identify any problems
> that I've missed so far.
>
> Any other thoughts?
>
> Thankyou,
>
> Dianne
>
> [Data/dti_erp/TBSS_ERP] dpat% rm -fr stats
> [Data/dti_erp/TBSS_ERP] dpat% tbss_3_postreg -S
> using pre-chosen target for registration
> transforming all FA images into MNI152 space
> ecc3131_FA_FAi
> ecc3393_FA_FAi
> ecc3472_FA_FAi
> ecc3515_FA_FAi
> ecc3559_FA_FAi
> ecc3591_FA_FAi
> ecc3647_FA_FAi
> ecc3717_FA_FAi
> ecc3740_FA_FAi
> ecc3824_FA_FAi
> ecc3835_FA_FAi
> ecc3888_FA_FAi
> ecc3936_FA_FAi
> ecc3956_FA_FAi
> merging all upsampled FA images into single 4D image
> creating valid mask and mean FA
> skeletonising mean FA
> now view mean_FA_skeleton to check whether the default threshold of
> 2000 needs changing, when running:
> tbss_4_prestats [threshold]
> [Data/dti_erp/TBSS_ERP] dpat% cd stats
> [dti_erp/TBSS_ERP/stats] dpat% fv
> [dti_erp/TBSS_ERP/stats] dpat% fslstats all_FA.nii.gz -V
> 0 0.000000
> [dti_erp/TBSS_ERP/stats] dpat% fslstats all_FA.nii.gz -M
> 0.000000
>
> On Jan 29, 2008 3:00 AM, Gwenaëlle DOUAUD <[log in to unmask]>
> wrote:
> Hi Dianne,
>
> it seems that at least a part of tbss_3_postreg has
> run properly. It should be quick indeed as it only
> applies the non-linear transformations (that you can
> see in the 4D images *_to_target_nonlinear.nii.gz)
> calculated at the tbss_2 stage to the corresponding
> images (*_to_target_nonlinear_hr.nii .gz). Then, it
> concatenates all the *_hr.nii.gz images in the big 4D
> file called all_FA.nii.gz. If this one is really
> empty, this is where the process failed. Have you got
> any error message running tbss_3? Maybe try running it
> again to see if the problem remains, and if this is
> the case, I'll give you the few command lines that
> will allow you to finish the tbss_3 stage and continue
> to the tbss_4 one.
>
> Hope this helps,
> Gwenaelle
>
>
> --
> Dianne Patterson, Ph.D.
> [log in to unmask]
> ERP Lab
> University of Arizona
> 621-3256 (Office)
---------------------------------------------------------------------------
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
---------------------------------------------------------------------------
|