[Freesurfer] Filipping issue

Dear List,

we are starting the analysis of cortical thickness for a group of patients who have only one side affected. The majority of patients have the left hemisphere affected, but some patients have the right hemisphere affected. In order to do a group analysis, we would like to flip the images of some patients, so that the affected side is left for all subjects.

Which is the best way to do this:

1) flip the 001.mgz volume and perform the entire recon-all pipeline with flipped images;

or

2) perform the recon-all pipeline with the non-flipped volumes and flip only the final surfaces for statistical purposes?

Thank you for any suggestion

Kind regards

Paola