The better skullstrip in AFNI

For most users of imaging software, we use the skull stripping program included in our distribution.  If you’re an AFNI user, you probably use 3dSkullStrip, if you use FSL then it’s BET, and of course Freesurfer has their own in recon-step 1.  A few years ago, I advocated for using Optibet and even wrote an entirely AFNI based version of the pipeline.

If you want to get all of the benefits of an Optibet like pipeline as well as doing your whole-brain normalization via a combination of linear and nonlinear warps, look no further than the “recently” introduced script in the AFNI distribution @SSwarper.  This program takes your original T1 anatomical image and does a robust skull stripping, as well as normalizes to the MNI152 brain via both a linear and nonlinear warp.  It even  generates helpful output images to double check the output like the one below!

@SSwarper \
-base MNI152_2009_template_SSW.nii.gz \
-input anatomical_scan.nii.gz \
-subid Sub123

So check it out, it’ll also make your afni_proc.py commands even more efficient, because you can feed the output to the AFNI superscript.  Helpful if you tend to run afni_proc.py with multiple options, you won’t have to rerun the normalization step over and over again (which can cost up to about an hour each time)!

Comments are closed.