Hi Kazu
The error that occurs while reading the structural is likely caused by a program failing to write the volume to disk. For example, this may happen if you stopped BET while it was trying to write the output.
You need to fix this before you can use this volume for further processing. E.g. by going back to the raw image, and re-betting etc.
Re registration: If you use the FDT GUI for registration, it will not, by default, produce the resampled volumes. You can use the transforms (in the xfms directory) to produce the transformed images using the GUI (in the main FSL GUI, click on MISC, then "Apply FLIRT transform")
Cheers,
Saad.
On 15 Sep 2010, at 00:30, Kazu Tanji wrote:
> Thanks Saad and Stam,
>
> When I tried to use main structural image I got the following error message:
>
> ERRORS: ++ WARNING:
> nifti_read_buffer(/data/DTI/KS-dicom_new.bedpostX/DBIEX_FSL_6_1_brain.nii.gz):
> data bytes needed = 167772160
> data bytes input = 158325424
> number missing =9446736
>
> Could you tell how I can check what is wrong with the file?
>
> For the meantime I used standard space for Registration.
>
> Regarding Saad's instruction, I don't see a transformed image created after Registration. Should it be created in the bedpostx directory? I only got transfer matrix files in xfms directory.
>
> Best,
> Kazu
>
>> Also, you need to use something like an MPRAGE as the Main Structural image, rather than the EPI b0 image, which is of lower quality.
>
>> Cheers,
>> Stam
>
>
> On 14 Sep 2010, at 09:14, Saad Jbabdi wrote:
>
>> Hi Kazu
>> You may have passed the wrong FLIRT transformation to probtrackx.
>> Please make sure that:
>> 1. The transformation from b0 to struct is correct (overlay the result on the structural in fslview)
>> 2. You pass to probtrackx the _inverse_ of the above transform (i.e. from struct to b0)
>> 3. Your seed/target masks are all in structural space
>>
>> Cheers,
>> Saad.
>>
>>
>> On 14 Sep 2010, at 06:28, Kazu Tanji wrote:
>>
>>> Hi list,
>>>
>>> I found the result of PROBTRACKX, fdt_paths.nii.gz, to be apparently misaligned.
>>> For Registration I used skullstripped b0 image as the Main structural image.
>>> I don't know if this is relevant but before that, I noticed BEDPOSTX had finished too early like in 6 hours.
>>> Is there a way to check what is going wrong?
>>>
>>> Best,
>>> Kazu
>>>
>>
>> --
>> Saad Jbabdi
>> University of Oxford, FMRIB Centre
>>
>> JR Hospital, Headington, OX3 9DU, UK
>> (+44)1865-222466 (fax 717)
>> www.fmrib.ox.ac.uk/~saad
>>
>
--
Saad Jbabdi
University of Oxford, FMRIB Centre
JR Hospital, Headington, OX3 9DU, UK
(+44)1865-222466 (fax 717)
www.fmrib.ox.ac.uk/~saad
|