External Email - Use Caution
Dear Manoj,
in case you want to rely on tools provided with FreeSurfer only, N3 can produce good results even for 7T data. However, in my experience, you need to supply a brainmask to N3 to achieve very similar results compared to SPM. Otherwise, especially around the temporal lobes the inhomogeneity will not be corrected sufficiently.
Tweaking recon-all for this is possible, but is slightly more complicated than changing the parameters of mri_nu_correct only. Furthermore, the procedure may change in case you use higher resolution 7T data.
Apart from that SPM, N3 as well as N4 of ANTs (also using a brainmask) will all produce similar results. I mostly use SPM for pre-processing my 7T data as it is the easiest to use in my processing pipeline.
Best,
Falk
________________________________ Von: freesurfer-bounces@nmr.mgh.harvard.edu freesurfer-bounces@nmr.mgh.harvard.edu im Auftrag von Manoj Saranathan manojsar@radiology.arizona.edu Gesendet: Montag, 6. Januar 2020 23:51:23 An: freesurfer@nmr.mgh.harvard.edu Betreff: [Freesurfer] Freesurfer 7T MPRAGE
External Email - Use Caution
Hi,
I am trying to process 7T MPRAGE data (not MP2RAGE) using recon-all. In previous threads, a few people have suggested using SPM to correct for bias first and then running recon-all. Alternatively I saw a thread which suggests modifying the mri_nu_correct command
https://www.mail-archive.com/freesurfer@nmr.mgh.harvard.edu/msg61913.html
Can I edit the recon-all to run this alternative command and simply run recon-all? The threads are all really old and SPM is now SPM12 which integrates things. Is this modified bias correction inside FS as good as SPM ? I am mainly interested in thalamic segmentation.
Are there others with the same issue?
Thanks
manoj
Manoj Saranathan Associate Professor, Dept. of Medical Imaging University of Arizona Ph: 520-626-6531