Hi,
The correct interpretation of a nifti image containing
the scale and slope is that the two images before
and after running it through FSL should have exactly
the same intensity range. That is, the original one should
have had the voxel values multiplied by the non-unity
slope, while the other one has a unity slope but should
contain values equal to the old values multiplied by
the old slope. Any software which behaves differently
in the two different cases is not really interpretting
the images correctly.
Hope this makes sense.
All the best,
Mark
On 2 Oct 2009, at 12:52, Rolf Heckemann wrote:
> Thanks Mark
>
> I processed a registration target image with fslswapdim, and the
> resulting image had an intensity ranging from 0 to 10^6. The IRTK
> non-rigid registration behaved strangely when I registered a source
> with an intensity range 0-500 with this target . I can work around
> the problem by dividing the target image by the original scale slope,
> so it's all right.
>
> If you're saying it's a feature rather than a bug, that's okay. I
> just never would have expected this behaviour.
>
> Best wishes
>
> Rolf
>
>
> On Fri, Oct 02, 2009 at 12:17:05PM +0100, Mark Jenkinson wrote:
>> That is what we have written it to do.
>> All FSL programs will do this.
>> Effectively we make our writing interface easier and
>> hence write out files which are equivalent to the input
>> files, but may be stored slightly differently, as is the
>> case here. Is it causing a problem with some software?
>>
>> All the best,
>> Mark
>>
>>
>> On 2 Oct 2009, at 11:32, Rolf Heckemann wrote:
>>
>>> When I use fslswapdim to convert a NIfTI file that has a scale slope
>>> greater than one, the resulting file has a scale slope of one and
>>> the
>>> original voxel values have been multiplied by the scale slope. Is
>>> this what it is supposed to do? I would have expected the resulting
>>> file to have the same scale slope as the original.
>>>
>>> Regards
>>>
>>> Rolf
>
>
> --
> Rolf A Heckemann, MD PhD
> Médecin chercheur
> Fondation Neurodis
> CERMEP - Imagerie du Vivant
> Hôpital Neurologique Pierre Wertheimer
> 59 Boulevard Pinel
> 69003 Lyon
>
>
>
> 1254483636
>
|