We get a lot of variable results with FLAIR. Sometimes it helps, sometimes it hurts. FLAIR scans themselves can be very variable across sites. If it is looking like it is hurting your results, then just run without it.

On 3/12/2021 11:07 AM, Xiaojiang Yang wrote:

External Email - Use Caution

Dear FS developers,

I have run FS’s recon-all on many cases, with and without -FLAIR options. I found the pial surface results (?h.pial) I obtained without -FLAIR option fit to the original image (subjected/mri/orig.mgz) much better than those with -FLAIR option, at least from my visual judgment. See MailScanner has detected a possible fraud attempt from "secure-web.cisco.com" claiming to be https://imgur.com/a/wrY0HTY for one example.

I know -FLAIR option was introduced to improve the quality of pial surface segmentation. Now for many cases, my intuition is that -FLAIR option decreases the pial quality. My questions are:

Is this intuition correct?

Should I trust the pial surface that’s obtained with or without -FLAIR option in general?

In addition, when I use the -FLAIR option, I sometimes get failed when running recon-all. After I did some manual edits to the brainmask image, and re-run the recon-all:

recon-all -autorecon-pial -subjid <subjectID>

I can get recon-all successful. But if I run recon-all without -FLAIR option, I can get successful results without needing to manually edit the brainmask image.

So my last question is:

Should I just run recon-all without -FLAIR option, or, with -FLAIR option + manual edit + re-run with -autorecon-pial? Which is better?

Thank you very much!

John
_______________________________________________
Freesurfer mailing list
Freesurfer@nmr.mgh.harvard.edu
https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer