I used mri_convert to resample my norm.mgz to 0.25 mm iso nii.gz, and then again back up to 1mm isotropic space. The resultant image shifted up a voxel from the original converted straght to nii.gz. This points to the resampling procedure, and not the conversion to different file format. Suggestions?<br>

<br>Note: I wanted to hand trace the hippocampus in the norm.mgz, and upsampled the norm.mgz to trace on. I traced it, and then downsampled to 1mm to compare with Freesurfer segmentation, but found it shifted.<br><div><font color="#999999">Josh</font><br>


</div><br>