External Email - Use Caution
Hi FreeSurfer team,
I’ve been using the Infant FreeSurfer pipeline on 3DT1 sequences for few weeks now on infants from 0.4 to 5.4 years-old.
I know Infant FreeSurfer is recommended for 0-2 years old infant but standard FreeSurfer was failing due to many lacunar infarcts in the white matter. So I tried Infant FreeSurfer with an age of 2 years old even for 2 to 5.4 years-old subjects and results look good. I know standard FreeSurfer has been tested on infant from 4.2 years-old, so what would be your recommendation for subjects from 2 to 4.2 years old? Have you already run Infant FreeSurfer on subjects older than 2 years-old? Do you recommend it?
I also have some processing issues (see below) with my actual data: 1. Pixels with 0 value => probably out of the brainmask so the pixels are set to 0 after masking So, sometimes lateral ventricles are miss-segmented (see "pixel_0_brainmask" attachment) 2. Thalamus region is missing from the aseg stat export (always equal to 0 mm3), do you have a fix? Should I fix a LUT file? 3. Would it be possible to get the non-skullstripped N3 corrected 3D T1 in the MNI space (3D iso)? I was not able to find the registered 3DT1 to re-apply the N3 correction, I only found the input 3DT1 (mprage.nii.gz) in its native space. 4. Do you have any workaround for large ventricles segmentation issue (see "large_ventricles" attachment)? Standard FreeSurfer has a -bigventricle option, do you have a similar approach?
And some additional questions: • brainstem seems to be split into multiple sub-regions, is it possible get the entire brainstem volume in the aseg stat export? • I saw that corpus callosum is not present in the “aseg” segmentation, so is this region out of the Infant FreeSurfer atlases? • Is it possible to get the mean cortical thickness for the labels you are providing?
Many thanks as I know there are plenty of questions here :) Best, Florent
