Dear Paul,
I haven't yet thought about this scenario. Even if images are initially affine registered I would recommend to use TIV as nuisance parameter. There might be still some variance that can be explained by TIV and you will be on the safe side to correct for it.
Best,
Christian
On Thu, 11 Aug 2016 11:12:49 +0800, KUN HSIEN CHOU <[log in to unmask]> wrote:
>Dear everyone
>
>This discussion list is very valueable for me. After reading the above
>discussions and related CAT12 manual , I think there are two main
>processing pipeline for CAT12 based customized DARTEL template. I have a
>question about how to adjust TIV in one of the VBM processing pipeline.
>
>For the VBM processing pipeline #1 , I summarilized the preprocessing steps
>in the manual and the above discussions
>==
>1. Run CAT12 segmentation and obtain the affine aligned GM and WM tissue
>segments for each individuals.
>2. Run SPM12 DARTEL tools (create template) to generate the
>"study-specific" DARTEL template using affine-aligned GM&WM segments as
>inputs.
>3. Run SPM12 DARTEL tools (Normalize to MNI space) to create smooth
>modulated GM tissue segments (using affine-aligned GM segments as input)
>4. Use these output smooth modulated GM segments for the follwing
>vxoel-wised statistical analsysis
>==
>For this pipeline , do we need to put TIV as covariate of non-interest in
>the design matrix ? Does this pipeline create "affine+non-linear" or
>"non-linear only" modulated GM tissue segments ?
>
>Best
>
>Paul
>
>2016-08-11 3:08 GMT+08:00 MRI More <[log in to unmask]>:
>
>> "You really only have to repeat the segmentation module if you think that
>> the first affine registration is going to differ to a relevant part from an
>> affine registration onto the study-specific template."
>> As it is rather unlikely that you are able to generate good
>> age-appropriate TPMs for six tissue types "on the fly" without additional
>> editing. Thus you would probably only be able to specifiy a different T1
>> image (and a different brain mask). This might be useful nonetheless, as it
>> will define the starting point of the Dartel registration. However, to
>> increase the registration accuracy overall I would assume the settings
>> during "Run Dartel (create Templates)" to be especially important, e.g.
>> some additional iterations, cycles, ..., and they might be the main
>> limiting factor as long as the image used for the initial affine
>> registration is not very different.
>>
>
|