A colleague asked me to look at some T1s for possible Freesurfer analyses. Unfortunately, several subjects do not have whole brain acquisitions (I know, quite unexpected and disappointing). Would it be possible to correct volumes of subcortical structures (which are covered by the acquisition) for ICV in the absence of whole brain coverage? The T1s were acquired in the coronal plane...so typically brain tissue is missing in either the anterior and/or posterior regions.
I reflexively think that whole brain coverage is required for accurate ICV calculated by Freesurfer...but I'm not positive.
Many Thanks for any thoughts you might have, Jenifer
Hi Jenifer,
how much is missing? If a substantial amount I would think things would fail, including the segmentation itself, but we very rarely do this.
Bruce
On Tue, 17 Nov 2009, Juranek, Jenifer wrote:
A colleague asked me to look at some T1s for possible Freesurfer analyses. Unfortunately, several subjects do not have whole brain acquisitions (I know, quite unexpected and disappointing). Would it be possible to correct volumes of subcortical structures (which are covered by the acquisition) for ICV in the absence of whole brain coverage? The T1s were acquired in the coronal plane...so typically brain tissue is missing in either the anterior and/or posterior regions.
I reflexively think that whole brain coverage is required for accurate ICV calculated by Freesurfer...but I'm not positive.
Many Thanks for any thoughts you might have, Jenifer
Hi Bruce, I think you might underestimate the robustness of Freesurfer :) In case I'm being too picky, I've attached a couple of example screenshots.
I appreciate that anyone rarely bothers with processing MRIs without whole brain coverage, but I'm wondering if *anything* can be salvaged from this dataset with many instances of incomplete brain coverage. I have 2 screenshots from the same subject (axial & sagittal planes) containing the norm.mgz (**orig.mgz**) with the aseg.mgz overlay. This subject was processed in FSv405. This example is not the worst (nor the best) example of incomplete coverage....varies quite a bit across subjects. The aseg.mgz looks pretty good. Will the ICV value generated by Freesurfer be accurate for correcting subcortical aseg volumes by ICV? Are there any "rules of thumb" as to what is necessary (in terms of brain/skull elements) for proper ICV values?
Many Thanks for any comments, Jenifer
-----Original Message----- From: Bruce Fischl [mailto:fischl@nmr.mgh.harvard.edu] Sent: Tuesday, November 17, 2009 12:05 PM To: Juranek, Jenifer Cc: freesurfer@nmr.mgh.harvard.edu Subject: Re: [Freesurfer] ICV for incomplete whole brain coverage
Hi Jenifer,
how much is missing? If a substantial amount I would think things would fail, including the segmentation itself, but we very rarely do this.
Bruce
that does look pretty good. My guess iss that the eTIV is also fine for this volume as it's only a 12 parameter transform, but I could be wrong (since I already was about our ability to accurately label this type of acquisition!)
Bruce On Tue, 17 Nov 2009, Juranek, Jenifer wrote:
Hi Bruce, I think you might underestimate the robustness of Freesurfer :) In case I'm being too picky, I've attached a couple of example screenshots.
I appreciate that anyone rarely bothers with processing MRIs without whole brain coverage, but I'm wondering if *anything* can be salvaged from this dataset with many instances of incomplete brain coverage. I have 2 screenshots from the same subject (axial & sagittal planes) containing the norm.mgz (**orig.mgz**) with the aseg.mgz overlay. This subject was processed in FSv405. This example is not the worst (nor the best) example of incomplete coverage....varies quite a bit across subjects. The aseg.mgz looks pretty good. Will the ICV value generated by Freesurfer be accurate for correcting subcortical aseg volumes by ICV? Are there any "rules of thumb" as to what is necessary (in terms of brain/skull elements) for proper ICV values?
Many Thanks for any comments, Jenifer
-----Original Message----- From: Bruce Fischl [mailto:fischl@nmr.mgh.harvard.edu] Sent: Tuesday, November 17, 2009 12:05 PM To: Juranek, Jenifer Cc: freesurfer@nmr.mgh.harvard.edu Subject: Re: [Freesurfer] ICV for incomplete whole brain coverage
Hi Jenifer,
how much is missing? If a substantial amount I would think things would fail, including the segmentation itself, but we very rarely do this.
Bruce
freesurfer@nmr.mgh.harvard.edu
-
Bruce Fischl -
Juranek, Jenifer