Hi Graham,

the ICV (or eTIV) we use is from a clever idea that Denise Head and Randy Buckner had. They showed that the total intracranial volume could be accurately estimated from the jacobian of the talairach transform (to an atlas with a skull), so the eTIV is just linearly related to the det of the talairach_with_skull.lta file, with the constant embedded in mri_label_volume I think. Basically we did a linear regression of the det with some manual TIV measures to determine in, then tested it with jacknifing.

cheers, Bruce

On Mon, 24 Jul 2006, Graham Wideman wrote:

Folks:

A couple of weeks ago, Bruce noted that:

...

EM Registration with Skull is for generating ICV measures

... so I just got around to investigating where the ICV measures might be getting saved.

Bottom line is I don't see what file that ICV measure is getting saved into.

Details:

I tried to figure out which program/script does this step, and what files it produces which might capture ICV.

1. Which commands correspond to the "EM Registration with Skull" step? Clue:

wiki recon-all shows this step employs the <no>skull-lta step. Looking in recon-all script, this corresponds to the DoSkullLTA variable, and that in turn leads us to mri_em_register as the pertinent command.

2. Looking at docs for mri_em_register in the wiki shows 4 arguments, only

one of which is flagged. Looking at actual help output from mri_em_register shows about 50 flagged arguments, none described.

3. The pertinent bit of script looks like this:

set xopts = `fsr-getxopts mri_em_register $XOptsFile`; set cmd = (mri_em_register -skull -t transforms/talairach.lta \ $xopts nu_noneck.mgz $FREESURFER_HOME/average/$GCASkull \ transforms/talairach_with_skull.lta)

4. From all this, I don't see what output file mri_em_register might be

placing the ICV into.

Clues?

Thanks,

Graham

---

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