Nifti can only handle 32k elements in any dimension. This is a problem for surfaces when they have upwards of 100k vertices. For our average subject (fsaverage and derivatives), this problem is solved by reshaping the number of vertices to be 27307x6=163842. If you are working with fsaverage, this should happen automatically (but might not for older FS versions). You can try converting to nifti using mri_surf2surf which offers a "--reshape" option.
doug
Gabriel Go.Es. wrote:
Dear freesurfers
I want to convert an .mgh image to NIfTY format, but even when the conversion using "mri_convert" seems to be ok, when I try to open the nii, an error message appears "image doesn't appear to be a valid 2D or 3D image", there is any other way to convert an mgh image to nii?
In other cases when I tryed to convert other mgh's cointaining clusters extracted with matlab but saved with the same matrix header as "sig.mgh", FS says that nrows in volume exceeds 32768, and does not make the conversion? Any ideas?
Thanks in advanced, Gabriel.
Freesurfer mailing list Freesurfer@nmr.mgh.harvard.edu https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer