Dear Chris,

thank you for the reply and for the insight.

Ok I see. The problem is when I load the data into software for processing, I get voxel dimensions 1.99 x 1.44 x 1.44 (I need isotropic voxels for tractography). In my understanding the voxel dimensions shouldn't change with rotation. I'm quite new to neuroimaging though so I may be missing something.

The software runs in Matlab and uses NIfTI tools by Jimmy Shen to load NIfTI data.

Best regards

Samuel