Hi Bruce,

Thanks a lot for the info. Would it mean we have to completely reprocess the data processed with 5.2? Or at which point of the preprocessing pipeline we should step in to fix the problems of 5.2...? I just have finished to process about 200 data sets with 5.2...

Best wishes,

Luke

Am 03.05.2013 um 02:58 schrieb "Bruce Fischl" <fischl@nmr.mgh.harvard.edu>:

> Hi All,
>
> We have recently uncovered problems with version 5.2. Based on these
> problems, we are recommending that people stop using 5.2, mainly in
> regards to usage of the surface data.  A new version, 5.3, will be
> released within two weeks, and a beta version for each platform will be
> available within a couple days.
>
> There were two main problems that were uncovered. First, a change was made
> that was only meant to affect the white surface placement in
> high-resolution data sets, but unfortunately is being applied to all data
> sets. This causes the white surface to contract a small amount. This causes
> the surface area to be smaller and cortex to be measured as thicker
> relative to version 5.1. Visually, the surfaces do not necessarily look
> less accurate, but we feel that the systematic differences with previous
> versions are ill-advised without a clear reason for the change.
>
> Second, a change made to the placement of the pial surface in order to
> improve the positioning based on T2/FLAIR that caused it to extend too far
> in some data sets during the T1-based pial surface deformation. This
> unfortunately became the default behavior, which was not intended. None of
> the dozens of test data sets we ran displayed this problem. The behavior
> was reported by other users which led us to the bug.
>
> In general, we rigorously test each version. However, our systematic
> comparison of a new version with previous versions which would have caught
> these problems was conducted prior to these changes. We are now tightening
> and regimenting our testing procedures to ensure this doesnt happen again.
> Note that there are no problems with the volumetric segmentation (aseg)
> in version 5.2.
>
> We realize that this will cause a substantial inconvenience for users. We
> have isolated and fixed the problems and have run two rounds of alpha
> testing and have one more to go, after which we will release a beta version
> in a couple of days to anyone that wants it.
>
> sorry for the problems.
>
> Bruce
> _______________________________________________
> 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.
>