Hi Elisabeth

we have found that higher voxel resolution doesn't necessarily help, as it starts to reveal inhomogeneities in the gray matter due to laminar structure, so 1mm isotropic is probably best. One approach we have used is to acquire 1mm data, then higher resolution in a specific part of cortex, although this can be difficult to analyze due to differential distortions between the scans.

cheers Bruce

On Mon, 9 May 2011, Wenger, Elisabeth wrote:

Dear Freesurfer experts,

I am currently planning a study in which I want to investigate changes in gray matter due to training on structural MR images. I heard some rumors that although FreeSurfer can of course process images with different voxel sizes, it always somehow "transforms" the voxel size to 1 x 1 x 1, irrespective of the originally obtained image resolution. I've been wondering if something has changed in this regard, that is, whether FreeSurfer is now capable of maintaining a higher resolution (eg, 0.8 mm) throughout all the analysis steps? As I am looking for very very tiny changes I would like to acquire images with 0.8mm resolution. But if FreeSurfer has to convert these images with the help of some estimation processes at some point of the reconstruction, I would rather like to acquire 1 x 1 x 1 images in the first place.

Thank you so much for your help!

All the best, Elisabeth