The reason is that the spatially normalized GM are written at a lower resolution than the original images. If the original resolution is 1mm and the resolution of the spatially normalized images is 2mm, then the accuracy of the volume estimates is likely to decrease because a lot of the information is missed out. I therefore chose to smooth the data slightly before resampling it to a lower resolution. As the VBM procedure is usually to smooth the data by about 12mm, I figured that this tiny bit of smoothing should be a perfectly reasonable thing to do.
Best regards,
-John
-----Original Message-----
From: SPM (Statistical Parametric Mapping) [mailto:[log in to unmask]] On Behalf Of Lemaitre, Herve (NIH/NIMH) [F]
Sent: Friday, July 06, 2007 3:47 PM
To: [log in to unmask]
Subject: Re: [SPM] VBM and smoothing
Hi Marko,
I discussed with Christian Gaser about that and he told me there was a smoothing and he removed it. I would prefer to use images without pre-smoothing but I would like to know if there is a specific reason to that in the original code.
Best
> Hi Herve,
>
>> I looked for an answer in the SPM list but i did not find it. I'm
>> using the last Gaser's VBM toolbox (version 1.10b). When I looked the
>> output images, they seemed to be less blurred than with the original
>> Segment button or the previous VBM toolbox. It seems that the Unified
>> segmentation adds a extra-smoothing step at the end of the procedure
>> while in the new Gaser's toolbox this step is removed.
>> Do we know the usefulness of this extra-smoothing?
>
>I would be surprised if there was an explicit smoothing step there... I would rather think that one of Christian's enhancements produces segmentation results that are a bit more crisp. This could be the
>no-priors option or the application of the hidden markov random field.
>Does this also occur if you use the exact same parameters?
>
>Best,
>Marko
>--
>=====================================================================
>Marko Wilke (Dr.med./M.D.)
> [log in to unmask]
>
>Universitäts-Kinderklinik University Children's Hospital
>Abt. III (Neuropädiatrie) Dept. III (Pediatric neurology)
> Hoppe-Seyler-Str. 1, D - 72076 Tübingen
>Tel.: (+49) 07071 29-83416 Fax: (+49) 07071 29-5473
>=====================================================================
|