I think these errors occur because occasionally an unfortunate confluence of circumstances leads an algorithm to make to an major error of judgement. The FreeSurfer process has randomness built into it so using a different random seed will avoid the bad circumstances and lead to a better outcome.
Peace, Matt. From: Antonin Skoch <[email protected]<mailto:[email protected]>> Date: Wednesday, November 9, 2016 at 2:52 PM To: Matt Glasser <[email protected]<mailto:[email protected]>>, "[email protected]<mailto:[email protected]>" <[email protected]<mailto:[email protected]>> Subject: Re: [HCP-Users] Pial surface misplacement Dear Matt, thank you for the suggestion! I tried to re-run FreeSurferPipeline.sh from scratch with different random seed and indeed the error disappeared. Could you comment on what precisely can be the reason of this error and why the re-run helped? Is it due to slightly different geometrical configuration of aseg which is dependent on value of initial seed? I am attaching screenshots with both runs with aseg overlaid. Blue surface is pial.postT2.pass1, yellow is pial.preT2.pass2. The blue surfaces are almost identical. What part of pipeline do you suggest to re-run in this case? I suppose the re-run from scratch had not been necessary to fix this issue... Is here anything I can do to avoid such type of issue in the future? Regards, Antonin From: "Glasser, Matthew" <[email protected]<mailto:[email protected]>> To: Antonin Skoch <[email protected]<mailto:[email protected]>>, "[email protected]<mailto:[email protected]>" <[email protected]<mailto:[email protected]>> Sent: 11/9/2016 12:21 AM Subject: Re: [HCP-Users] Pial surface misplacement I would try running with a different random seed and see if this fixes this issue. Peace, Matt. From: <[email protected]<mailto:[email protected]>> on behalf of Antonin Skoch <[email protected]<mailto:[email protected]>> Date: Tuesday, November 8, 2016 at 5:19 PM To: "[email protected]<mailto:[email protected]>" <[email protected]<mailto:[email protected]>> Subject: [HCP-Users] Pial surface misplacement Dear experts, I am checking quality of the pial and white surfaces generated by HCP pipeline in our data (with acquisition parameters set according to the HCP recommendations). I found that in some spots the pial surface does not seem to reasonably follow the underlying anatomy - it cut too much of gray matter and reaches almost to the white surface. I tracked the problem to the step in FreeSurferHiResWhite.sh where the file ?h.preT2.pass2 is generated (see the attached screenshot, blue surface is the pial.postT2.pass1, yellow surface is pial.preT2.pass2). Could you please comment on where could be the problem and how to face it? I suspect that this kind of error could have profound effect on the results. My second inquiry is, are the source codes of mris_make_surfaces of HCP version of FreeSurfer available? Or, how is the difference in code between standard 5.3.0 and HCP version? Regards, Antonin Skoch _______________________________________________ HCP-Users mailing list [email protected]<mailto:[email protected]> http://lists.humanconnectome.org/mailman/listinfo/hcp-users ________________________________ The materials in this message are private and may contain Protected Healthcare Information or other information of a sensitive nature. If you are not the intended recipient, be advised that any unauthorized use, disclosure, copying or the taking of any action in reliance on the contents of this information is strictly prohibited. If you have received this email in error, please immediately notify the sender via telephone or return mail. ________________________________ The materials in this message are private and may contain Protected Healthcare Information or other information of a sensitive nature. If you are not the intended recipient, be advised that any unauthorized use, disclosure, copying or the taking of any action in reliance on the contents of this information is strictly prohibited. If you have received this email in error, please immediately notify the sender via telephone or return mail. _______________________________________________ HCP-Users mailing list [email protected] http://lists.humanconnectome.org/mailman/listinfo/hcp-users
