Dear FSL crew,
I'm performing a typical DTI group analysis. However, I have some subjects that are critically cut due to a bad acquisition process (bad luck!). Applying an explicit mask would be an option to limit the GLM to all voxels that exceed a concrete value in all subjects. This approach, however, would exclude some brain regions that are interesting for my analysis (i.e. we would exclude a region due to only few subjects cut, ignoring the information that all other subjects have in this region).
I'm wondering wether it's possible to do a more sofisticate approach. I mean, for example, for each voxel include all subjects that have information and exclude those that do not: a different size glm in each point. This way, we would not loose any information in any voxel.
Thanks in advance for your opinions,
Best.
|