Re: [Freesurfer] input anatomy orientation

This is expected, unfortunately. When you change the orientation, it changes the interpolation which changes things slightly. Those areas are all pretty small so the changes can appear to be large (though they are only a few percent).

doug

On 07/09/2014 01:42 PM, Kathy L. Pearson wrote:

I have run FreeSurfer 5.3 recon-all, getting different stats output, for these cases of anatomy input: (1) mprage DICOM series converted with dcm2nii (2) fslreorient2std (RL PA IS) on the dcm2nii output (3) mprage DICOM series converted with mri_convert

Differing example results include these from lh.BA.stats: case 1: BA1 1115 583 2017 2.511 0.650 0.175 0.155 66 9.1 BA2 3686 2392 5986 2.342 0.615 0.123 0.047 44 7.2

 case 2:
     BA1                                      1043    585   2060  2.569 0.637     0.181     0.226      169     4.8
     BA2                                      3569   2357   6064  2.397 0.612     0.123     0.050       38     5.9

 case 3:
     BA1                                      1071    558   1994  2.583 0.617     0.159     0.181       66     6.5
     BA2                                      3654   2364   6191  2.461 0.631     0.121     0.058       42     6.9

Is this as expected? Are there any options that I should apply to indicate that the orientation is RL PA IS if I use the same anatomy.nii.gz file in case #2 as prepared for FSL? Or must I always either use mri_convert to generate the input for recon-all? Or is there a command I should be running first to reset my RL PA IS image to a particular orientation required by recon-all?

Thanks,

Kathy Pearson

---

Freesurfer mailing list Freesurfer@nmr.mgh.harvard.edu https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer