Hi Michael,

looks like this defect has about 60K vertices in it, which is almost have a typical surface. Something is dramatically wrong. Look at the ?h.orig and see if it includes skull or cerebellum, or if there is some huge hole/handle in the ?h.orig.nofix.

cheers Bruce

On Wed, 26 May 2010, Michael Waskom wrote:

Hi, all,

We have a recon that is failing during topology correction. Here's the history or what's happened: --first run through the recon-all -all pipeline went fine --control points were added and pial edits were made for quality control --a second recon was run with --autorecon3-cp --autorecon3

During the second recon, the process gets stuck for a long time in topology correction and eventually crashes. We've tried it twice, and it crashes on the same defect.

Here's info from the bugr script:

Debian version: 5.0

Kernel info: Linux 2.6.28-18-generic x86_64

recon-all log is attached.

Any ideas? Thanks!

Mike