Hi Yezhe

 

It depends a lot on what your resolution, read out direction, bandwidth and field strength are. We generally fix these with T2 or FLASH scans at the same resolution/geometry/readout as the T1. Otherwise they are pretty tough as the dura is close enough to the gray matter that you can’t see any space between them

 

Cheers

Bruce

 

From: freesurfer-bounces@nmr.mgh.harvard.edu <freesurfer-bounces@nmr.mgh.harvard.edu> On Behalf Of LIN, Yezhe
Sent: Monday, November 30, 2020 5:10 PM
To: freesurfer <freesurfer@nmr.mgh.harvard.edu>
Subject: [Freesurfer] Freesurfer skullstripping error

 

        External Email - Use Caution        

Dear Freesurfer users,

 

Hope this finds you well.

I preprocessed our dataset with recon-all <subj> -s <subj> -all -qcache and found a number of our subjects had different levels of skullstrip errors. I tried fixing them by a couple of different approaches, such as -gcut and adjusted watershed threshold, but they both failed to fix the overlapping of dura mater (Fig1 and 2). I tried two methods on different subjects and received the same outcome. I wonder if you have any ideas to fix this error. At this point, for some who had less skullstripping, I can erase them manually and run -autorecon-pial or -make. But for the others with multiple locations of skullstripping (like Fig 3 and 4), do you have any ideas on how to approach to them at a systemic level? Also I found the skullstripping error in over half of our subjects, is it normal that we have a such proportion?

 

Thank you so much!

 

------------------

Best regards,

Yezhe

Yezhe Lin, MD MMed  (she/her/hers)

Postdoctoral Associate, Department of Psychiatry

University of Rochester Medical Center

yezhe.lin.bee@gmail.com

Twitter: @LinYezhe | LinkedIn | Researchgate