for 3T scans, try this modification to the nu_correct usage in recon-all. first, create a file called 'xopts' with this one line:
mri_nu_correct.mni --proto-iters 1000 --distance 50
then, run recon-all with these params:
recon-all -s subjid \ -nuintensitycor \ -nuiterations 1 \ -expert <path>/<to>/xopts
the new flags being '-nuiterations 1 -expert ./xopts'.
these params are taken from a new paper:
http://web.mysites.ntu.edu.sg/zvitali/publications/documents/N3_NI.pdf
also, mri_nu_correct.mni --help provides help info.
n.
On Mon, 2010-04-26 at 21:48 +0200, Leonardo Chesi wrote:
Dear all,
I’m Leonardo, a student of the University of Pisa. We study AD’s patients by MRI. In the last days I had have many problems with the segmentation because of the B0 inhomogeneity with a 8 channel phased array coil on a 3.0 T General Electric machine. In some articles I have found the nu_correct command that provide to normalize the image. I’m writing to you to ask if some of you know the optimal parameters to give input to nu_correct to normalize as well as it possible the image. I’ve try with the default parameters and my segmentation is already better than the previous.
Moreover I’d like to know if there is a manual of this command because on the FreeSurfer Wiki I can’t find detailed information.
I’m waiting for your corteous help.
Greetings
Leonardo
Freesurfer mailing list Freesurfer@nmr.mgh.harvard.edu https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer