Hi all, I'm posting a follow-up to my thread from November regarding "anterior temporal lobe problems", so that it can be captured in the FS archives.
Based on a suggestion from Bruce, we tried re-running (from scratch) using the -mprage flag in recon-all (since ours were MPRAGE acquisitions) as well as "-b 20 -n 5" as expert options for the mri_normalize command.
These changes substantially improved the quality of the white (and thence pial) surfaces in many of the subjects of our pediatric brain sample! Many subjects went from having moderate problems in the surfaces at the very tip of the temporal lobe to having no problems there, and other subjects went from having major surface errors throughout the anterior and inferior temporal lobe to more minor, isolated problems.
I did a very minimal investigation regarding the independent impact of the -mprage, -b 20, and -n 5 contributions to the improvement (i.e., in just one subject) and found that (in that one particular subject) the "- b 20" inclusion as an expert option for mri_normalize was sufficient to account for almost all the improvement. However, the -mprage flag by itself also yielded substantial improvement. The "-n 5" inclusion as an expert option didn't, by itself, yield any improvement.
Hope that helps. If you have problems with your surfaces in the anterior/inferior temporal lobe, you might want to try these modifications. And please post what you learn back to the list as to whether they were helpful in your particular case.
cheers, -MH
On Thu, 2011-11-03 at 17:22 -0400, Khoa Nguyen wrote:
Control points do a good job in fixing the problem for me. You just have to put a string of control points (not too many) on the white matter strips for every 2-3 slice.
-Khoa
On Thu, 3 Nov 2011, Louis Nicholas Vinke wrote:
Hi, Some other potential explanations for poor surfaces in this area are addressed here: https://surfer.nmr.mgh.harvard.edu/fswiki/UserContributions/FAQ#Q.Thesurface...
"This region tends to be furthest away from the coils receiving the MR signal, potentially late to myelinate in individuals under 20 years old, very thin white matter in general, and is in close proximity to regions of susceptibility."
I'm not aware of any flags which could begin to fix the surfaces in those areas automatically. -Louis
On Wed, 2 Nov 2011, Freesurfer Local Archive wrote:
Michael,
I had a similar problem with about 10% of pediatric datasets I used. In my case the problem also often included anterior and ventral surface of the frontal lobes. Much like you I looked at possible explanations and a medical physicist and a radiologist I worked with blamed inaccuracies entirely on chemical shift. Adding control points at least in temporal lobes generally provided only moderate improvements.
Greetings,
Jacek
On Wed, Nov 2, 2011 at 4:02 PM, Michael Harms mharms@conte.wustl.edu wrote:
Hi guys, We have a sample of 150+ pediatric brains that we pumped through FS 5.1, and in an easy majority of them the white/pial surface accuracy in the anterior temporal lobe is poor.
I know that this is an area where problems have frequently been reported. What is it about that area in particular that makes it difficult to get accurate surfaces? (Is the contrast and SNR in that area inherently worse?) And, short of a large time commitment of control point and wm.mgz editing, are there any flags/tweaks that we could attempt to fix the problem in a more automatic fashion?
thanks, -MH
-- Michael Harms, Ph.D.
Conte Center for the Neuroscience of Mental Disorders Washington University School of Medicine Department of Psychiatry, Box 8134 Renard Hospital, Room 6604 Tel: 314-747-6173 660 South Euclid Ave. Fax: 314-747-2182 St. Louis, MO 63110 Email: mharms@wustl.edu
Freesurfer mailing list Freesurfer@nmr.mgh.harvard.edu https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer
The information in this e-mail is intended only for the person to whom it is addressed. If you believe this e-mail was sent to you in error and the e-mail contains patient information, please contact the Partners Compliance HelpLine at http://www.partners.org/complianceline . If the e-mail was sent to you in error but does not contain patient information, please contact the sender and properly dispose of the e-mail.
Freesurfer mailing list Freesurfer@nmr.mgh.harvard.edu https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer