Print

Print
Hi Mark

In Normalise Estimate and Write the 3 parameters you mention seem ok to me.

Regarding your 2 questions, 

(1) Yes, you can use the _sn parameters to normalise the functional data of that subject. However, a lot of people in the SPM community are now moving towards DARTEL for normalisation (when you have good quality anatomical scans). See the SPM manual Chapter 26 for information on the DARTEL tools.

(2) No, Not quite. Coregistration does no warping to the image, it just moves it around (for example, to make sure that your anatomical image is well registered with the mean EPI volume in space for a given subject). Normalisation does a set of non-linear transformations to warp the image to best match the template (in the ICBM/mni space for example).

I recommend that you have a look at the example analyses in the SPM manual Chapters 28-30.

Hope this helps!

Ian





2014/1/12 Mark <[log in to unmask]>
Dear Ian,

I have to specify 3 items in Normalise & Write:

Source Image: ? (I input T1 image)
Images to write: ? (I input T1 image again)
Template image: T1.nii

Are my procedures correct?

It then output a wxxx.img file and I've checked it and I think this is the T1 image normalised to standard space. Right?

I also noticed 2 questions:

(1).
I noticed that it also output a xxx_sn.mat file. According to the manual, this file contains parameters that best register a source image (or series of source images) to match the template. Below is my understanding: If I want to normalise this subject's functional images (or other images) to standard template, I can select "Normalise (Write)" then in Data, Parameter file = xxx_sn.mat, Images to write = functional images. Am I correct?

(2).
The process is quite like "Coregistration (Est&Res)," in which the Source Image is moved to match the Reference Image (and a xxx_sn.mat is also produced). To normalise this subject's functional images to standard template, can I select "Coregistration (Est&Res)" and set Source Image as T1.nii and Reference Image as functional images?

Thanks

Mark