Nick, This is great -- you probably also noticed that the version of the transform used (without or with skull) can make a big difference in the computed scale factor.
Was variability assessed by comparing eTIV values computed from multiple MR sessions from the same subject? Or by comparison to manually traced ICV from a T2w scan? Could you provide some sense of what the actual variability (e.g., coefficient of variation) was for the old vs. new approaches? Regarding the accuracy of the new eTIV calculation: Since the scale factor is now going to be derived from a transform without the skull, do you have a sense to what degree atrophy and increased sulcal CSF will impact the values? That is, should the new measure be considered more an estimate of "TBV -- total brain volume" (defined as the total volume of everything interior to the outer pial surface, including ventricular CSF), rather than "intracranial volume" per se? thanks, Mike H. On Wed, 2009-02-11 at 18:41 -0500, Nick Schmansky wrote: > If you make use of the estimated Total Intracranial Volume (eTIV) > calculation found in the aseg.stats file of your subject data, then you > should be aware that we have recently found a more accurate way to make > this calculation. We have found that using the talairach.xfm transform > instead of the talairach_with_skull.lta transform, from which the > determinant is extracted and multiplied by a scale factor, results in > lower variability of the eTIV calc in our test set. > > This new way of calculating eTIV will appear in the next release of > Freesurfer. > > If you want to make use of this new calculation right away, you can > download a new version of mri_segstats from here: > > ftp://surfer.nmr.mgh.harvard.edu/pub/dist/freesurfer/fixes > > and run it using this command: > > mri_segstats --subject subjid --etiv-only > > and it will output the new eTIV. > > Nick > > > _______________________________________________ > Freesurfer mailing list > Freesurfer@nmr.mgh.harvard.edu > https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer _______________________________________________ Freesurfer mailing list Freesurfer@nmr.mgh.harvard.edu https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer
