Dear Julia,
for the initial spatial normalization and segmentation I would recommend to create a customized TPM using the Template-O-Matic Toolbox(use an age of 5.5 years for creating the TPM).
For the DARTEL normalization it is also recommended to create a customized DARTEL template using both time points. If you intend to use the CAT12 toolbox you should keep in mind that this longitudinal pipeline is optimized for rather small changes between the time points (e.g. learning effects, plasticity). However, you can give it a trial, but I would also try the Longitudinal Toolbox in SPM12 that can better deal with larger changes between the time points. In CAT12, there is an option to save affine registered segmentations of the average for all time points that can be used in order to create a customized DARTEL template.
Best,
Christian
On Thu, 20 Oct 2016 12:35:31 +0000, Binder, Julia <[log in to unmask]> wrote:
>Dear VBM-experts,
we plan to analyze structural MRI data of a longitudinal study with children (time point 1: 5 years old, time point 2: 6 years old). As there have not yet been published a lot of studies, we wonder whether anyone has experience with analyzing longitudinal data of children of this age range?
Our specific question concerns the preprocessing with a sample-specific template: Would you create a sample-specific template based on all scans at time point 1 and 2 or for each time point separately?
I would be very interested in any insights from you.
Thanks a lot in advance.
Cheers,
Julia
|