Hi Jenifer
sometime scanners will upsample by zero-padding on the scanner. This doesn't actually create any information, but some clinicians like it. That is, your underlying data may actually be close to 1mm iso, even though the in plane is apparently 0.48mm. You can probably tell this visually (if it is not terribly noisy). In that case highres is not worth it. I guess you can try it both ways and see Bruce
On Sun, 10 Jun 2018, Juranek, Jenifer wrote:
External Email - Use Caution
I offered to help a colleague with some Freesurfer analyses, but when I looked at the data they were processing, I found:
- highly anisotropic voxdims
- sub-mm in-plane resolution
Data were acquired on a 3T Siemens Tim Trio. I don’t think they were trying to deviate from the recommended 1x1x1mm specs, but somehow it looks like they did since the 001.mgz file does not conform to the FS standard. Output for mri_info on their 001.mgz file is listed below. They did not keep the file converted by mri_convert (nii or dicom) in their FS_sub directory, so I’m limited to the 001.mgz. I’m running FSv6 in my lab, is the hires flag appropriate in this situation? Any suggestions? Many Thanks, Jenifer
[orig]$ mri_info 001.mgz Volume information for 001.mgz type: MGH dimensions: 448 x 512 x 192 voxel sizes: 0.478516, 0.478516, 1.000000 type: SHORT (4) fov: 245.000 dof: 0 xstart: -107.2, xend: 107.2 ystart: -122.5, yend: 122.5 zstart: -96.0, zend: 96.0