Does anyone out there have a script for running longitudinal VBM analyses they would be willing to share? Something that links the steps below as previously suggested by Gwenaelle on the list.
1) determine the halfway transformation (as in Siena) between the two timepoints for each subject (a priori 6DOF)
2) register both timepoints to this halfway space, average them, then bet the average
3) realign the bet mask in each native space for both timepoints and then segment each image with fast
4) register both GM timepoints to the halfway space and then average them
5) fnirt the GM average from the halfway space of each subject to the MNI space
6) convolve the subsequent warpfield with the halfway .mat to obtain the final warpfield to apply to each native GM image
7) modulate and smooth