On Fri, 7 Oct 2005 16:03:31 +0200, Anja Ischebeck
<[log in to unmask]> wrote:
>Hi Stephen,
>
>my impression that SPM2 overregularizes less than SPM99
>(given the default values of both) stems from just practically
>comparing the normalization procedure of SPM99 and SPM2
>(using the EPI-template) in the case of a whole brain
>functional image with a great deal of signal loss in the
>frontal orbital cortex. This presents a similar problem,
>because the normalized EPI-Templates contains tissue in this area.
>In effect the brain normalized with SPM99 'grew' tissue into
>the non-existent area (approx an area of 5 cm!)
>This did not happen with SPM2 for whatever reason.
>
>I assumed that the reason for this is the new cut-off value option
>in SPM2 (the regularization option is present in both).
>But, of course, certainly more has changed between SPM99 and SPM2
>regarding normalization than just that.
Right. I'm guessing it has to do with the energy/penalty function used in
the normalization algorithm, but obviously John (Ashburner) is the one
who'd really know.
Cheers,
S
>
>Hope that clarifies the matter
>Best,
>
>Anja
>
>
>
>
>Dr. Anja Ischebeck
>Innsbruck Medical University
>Clinical Department of Neurology
>Anichstrasse 35
>A-6020 Innsbruck - Austria
>tel.: +43 (0) 512 504 23661
>
>>>> "Stephen J. Fromm" <[log in to unmask]> 07.10.2005 15:41 >>>
>On Fri, 7 Oct 2005 10:03:53 +0200, Anja Ischebeck
><[log in to unmask]> wrote:
>
>>Hi Philipp,
>>
>>Overnormalization is less likely in SPM2 than in SPM99.
>>There is a cutoff-value in the defaults (which you could
>>change yourself to an even more conservative value)
>>that prevents the normalized brain to extend beyond
>>the area covered by the scans.
>>Using the EPI template should therefore work ok.
>
>What cutoff is that? AFAICT there's a similar cutoff in SPM99. It was
>
>just named differently (number of basis functions instead of size in
>mm).
>
>My impression is that if part of a source brain is missing,
>normalization
>algorithms will invariably try to "grow it" towards the edge of the
>template. (One way to try to deal with that is to change the
>regularization, as someone just posted.)
>
>Cheers,
>
>S
>
>>However, using the EPI-template can introduce
>>spatial insecurities of approx. 2 cm! We tested both
>>ways of normalization (using T1 or EPI template)
>>with a tumor data set - with this quite alarming result,
>>especially if you are interested in small brain areas
>>(hippocampus and the like).
>>
>>If coregistration fails (it does in 50% of all cases with MI
>>at least for us) you can work around this in three ways:
>>
>>Use a deskulled anatomical
>>(but don't use this image for normalisation!
>>so copy the .mat file gained and rename it so it applies
>>to the original anatomy)
>>
>>Use the coregistration in SPM99 (the segmentation approach
>>is worth (several) tries, if MI fails. If MI fails, it will fail
>>consistently
>>whatever the starting parameters and default options -
>>at least to my experience. When Segmentation fails you get at least
>>different results for different starting parameters (.mat file).
>>
>>If you get a near but not perfect match, e.g. the funcs 2mm lower
>>than the anatomicals, use display to enter the displacement value
>>by hand and reorient the anatomical.
>>
>>Hope that helps,
>>Best,
>>
>>Anja
>>
>>
>>
>>
>>
>>
>>Dr. Anja Ischebeck
>>Innsbruck Medical University
>>Clinical Department of Neurology
>>Anichstrasse 35
>>A-6020 Innsbruck - Austria
>>tel.: +43 (0) 512 504 23661
>>
>>>>> Philipp Saemann <[log in to unmask]> 07.10.2005 01:49 >>>
>>Hi,
>>
>>for an fMRI series which does not cover the whole brain (e. g. 20
>>slices 4 mm
>>thick), how is spatial normalisation best achieved?
>>
>>Can object masking (be recommended, e. g. a binary mask of the
>volume)
>>to keep
>>the influence of the missing parts and the border zones low ?
>>
>>Should the starting position be optimized manually (SPM2)?
>>
>>(Coregistering on T1 did not work properly using MI, otherwise we
>would
>>have
>>used this option)
>>
>>Grateful for any hints on this,
>>
>>Philipp
>>=========================================================================
>=========================================================================
|