On 08/06/2014 11:40 AM, Krieger, Donald N. wrote:

mri_cor2label does work for extracting vectors with FreesurferColorLUT.txt label numbers from .mgz volumes.

The documentation at http://surfer.nmr.mgh.harvard.edu/fswiki/mri_vol2label http://surfer.nmr.mgh.harvard.edu/fswiki/mri_vol2label states:

--id labelid

0-255 value in COR volume

That is probably old documentation (try running it with --help)

The attached pdf shows the number of vectors which were extracted from each of 5 volumes for labels 1-255.

You can see that volumes aseg, aparc+aseg, and wmparc yield nearly identical results. I ran several file pairs through spot checks and the vectors were identical in the ones I checked.

Volume filled.mgz contains vectors for only 3 labels while volume T1.mgz contained vectors for all of them and in most cases, the numbers were higher.

Questions:

1.Can mri_cor2label handle labelIDs > 255?

yes

2.Is there a tool determining which labelIDs are represented in a .mgz volume?

mri_segstats will do it (see the --sum output)

3.Are there other volumes which I should be using for vector extraction for these labels.

I'm not sure what labels you want.

4.Is there a way to extract labels with ids > 255? I assume that mri_cor2label is an older tool and the new approach is generate .annot files and then use mri_annotation2label.

mri_cor2label

5.How are .annot files created? I have been unable to find this even though it must be staring me in the face.

mris_ca_label during recon-all

Thanks,

Don

---

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