Nicolas Cherbuin wrote:
Hi,
Although the documentation for fs is fantasic there are a few questions I have not been able to answer to my satisfaction. I hope somebody will have some patience left to help me out :-)
- When using mris_preproc we must provide an fsgd file, it is not clear to me if and how the information in the fsgd file is used in this step. Since this file is required I assume that it is not possible to preprocess the data once with a generic fsgd file and then perform glmfit with different fsgd files but just in case I thought I'd check.
Actually, it is not required. It is only used to get the list of subjects to use. By using the same fsgd file for preproc and analysis, you assure that the order of the data and the subjects is the same. But you can certainly use one generic fsgd and then analyze the result with many different ones.
- Next to the FDR rate in the configuration overlay window there is a "only marked" tick box. How is this meant to be used?
There should be a text box there to put the FDR.
- I have searched the mailing list and the web but I am still a bit confused about what strategy works best in practice for mutliple comparison correction. My experience so far is that if I use FDR most results disappear....
This is a big question, and there is not a great answer. FDR may be more conservative than a cluster-based technique. For more on FDR, I'd recommend seeing Tom Nichols' web site www.sph.umich.edu/~nichols/FDR/
doug
Thanks
Nic
Freesurfer mailing list Freesurfer@nmr.mgh.harvard.edu https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer