Dear Jon,
You can overlay the path distributions on any image in the same space -
for paths in diffusion space you may want to overlay them on the
nodif_brain image. Alternatively, you can transform that subject's
structural scan (if you have it) into diffusion space by applying the
str2diff.mat transform (if you have generated it) and use that as a
background image.
You mention that some of your tractography results look unlikely. You
should be able to tell reasonbly easily whether some images are back to
front - seed a number of single voxels in clear white matter tracts,
e.g., internal capsule, SLF, corpus callosum etc, and check that outputs
are plausible. If you are still concerned then please put an example of
an output on the web somewhere for us to look at.
Heidi
Jon Clayden wrote:
> Hi Stephen,
>
> The platform is Red Hat Linux 9 (x86). The output data look plausible,
> except for the fact that superimposing them onto an FA map generated by
> dtifit sometimes shows bright areas in CSF, as I mentioned. Is it valid
> to overlay these two images, and if not, what should I be putting
> "under" the probtrack results in fslview?
>
> Thanks,
> Jon
>
> On 6 Nov 2004, at 10:04, Stephen Smith wrote:
>
>> Hi Jon.
>>
>> Don't worry about the -ve pixdims - that's something we'll tidy up in
>> a future release but they should always not make a difference anyway.
>>
>> The analysis could well be that quick if your data is lower
>> resolution/field-of-view than we used for the example timings in the
>> manual.
>>
>> The error message on completion looks odd, particularly as the bedpost
>> script is only 293 lines long. What platform are you using? It does
>> look
>> as if it probably did finish fine - I assume that the output data
>> looked
>> ok.
>>
>> Cheers.
>>
>>
>>
>> On Wed, 3 Nov 2004, Jon Clayden wrote:
>>
>>> Hi,
>>>
>>> I have a couple of questions that are bothering me and don't seem to
>>> be
>>> answered by the FAQ.
>>>
>>> First, I've noticed that running avwhd on some of the Analyze files
>>> generated by various stages of the FDT pipeline produces a seemingly
>>> arbitrary mixture of positive and negative "pixdim1" values. (See, for
>>> example, the bedpost_datacheck output below.) Is this expected; and if
>>> so, what is happening? When I overlay output from probtrack with an FA
>>> map from the same data the tractography results look rather unlikely
>>> in
>>> parts, occasionally straying into CSF, so I wondered if one of the
>>> images might be back-to-front with respect to the other.
>>>
>>> Secondly, when I run bedpost it processes the data for about 12 hours,
>>> which seems suspiciously short considering the spec of the box it's
>>> running on (Red Hat 9 derivative, 2GHz or so, 512Mb RAM), and then
>>> finishes with output like the following:
>>>
>>> [...]
>>> 40 slices processed
>>> 41 slices processed
>>> 41 slices processed
>>> 42 slices processed
>>> 43 slices processed
>>> 44 slices processed
>>> 45 slices processed
>>> Merging outputs into 4D files
>>> DONE
>>> /disk/home/cornell/s0343526/fsl/bin/bedpost: line 294: syntax error:
>>> unexpected end of file
>>>
>>> Should it be finishing at this point, or is something really wrong?
>>>
>>> Thanks,
>>> Jon
>>>
>>>
>>> bedpost_datacheck output:
>>>
>>> ./data
>>> dim1 128
>>> dim2 128
>>> dim3 48
>>> dim4 52
>>> datatype 4
>>> pixdim1 -1.7187500000
>>> pixdim2 1.7187500000
>>> pixdim3 2.7999999523
>>> pixdim4 1.0000000000
>>> cal_max 0.0000
>>> cal_min 0.0000
>>> glmax 0
>>> glmin 0
>>> origin1 0
>>> origin2 0
>>> origin3 0
>>> file_type ANALYZE-7.5
>>>
>>> ./nodif
>>> dim1 128
>>> dim2 128
>>> dim3 48
>>> dim4 1
>>> datatype 4
>>> pixdim1 1.7187500000
>>> pixdim2 1.7187500000
>>> pixdim3 2.7999999523
>>> pixdim4 1.0000000000
>>> cal_max 1385.0000
>>> cal_min 0.0000
>>> glmax 0
>>> glmin 0
>>> origin1 0
>>> origin2 0
>>> origin3 0
>>> file_type ANALYZE-7.5
>>>
>>> ./nodif_brain_mask
>>> dim1 128
>>> dim2 128
>>> dim3 48
>>> dim4 1
>>> datatype 4
>>> pixdim1 -1.7187500000
>>> pixdim2 1.7187500000
>>> pixdim3 2.7999999523
>>> pixdim4 1.0000000000
>>> cal_max 1.0000
>>> cal_min 0.0000
>>> glmax 0
>>> glmin 0
>>> origin1 0
>>> origin2 0
>>> origin3 0
>>> file_type ANALYZE-7.5
>>>
>>> num lines in ./bvals
>>> 1
>>> num words in ./bvals
>>> 52
>>> num lines in ./bvecs
>>> 3
>>> num words in ./bvecs
>>> 156
>>>
>>
>> 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
>
--
Dr Heidi Johansen-Berg
Wellcome Trust Training Fellow
Oxford Centre for Functional Magnetic Resonance Imaging of the Brain
John Radcliffe Hospital
Headington
Oxford OX3 9DU
http://www.fmrib.ox.ac.uk/~heidi
Tel: 01865 222782
Fax: 01865 222717
email: [log in to unmask]
|