Hello gentle readers and experts of FreeSurfer,
I am using the FS software package to calculate the gyrification index for premature baby brain. Of course in order to get to this step I do need to run the previous autorecon steps to obtain the pial surface.
So far I've run the first sequence for one of my test subjects, recon-all -autorecon1, and encountered no errors. For the second step autorecon2 I have decided to run every subsequent step, firstly, because if I understand correctly from the wiki-webpage the flag -noaseg is necessary if the subjects cannot use an aseg, like baby brain. Secondly, due to the poor T1,T2 intensity contrast in my baby-scans the FS default segmentation step was not performed correctly (large portions of white matter were not counted). So I include my own segmented white matter volume. To include my WM volume I have used "mri_convert -c" to convert analyze data into .mgz format.
The steps from the second sequence that I ran without getting any errors are: "-gcareg", "-canorm", "-careg", "-careginv", "- rmneck", "-skull-lta", "-normalization2 -noaseg", "-maskbfs" .
At the step "-fill -noaseg" the errors are:
min_slice = -1, min_area = 65536 mri_fill: could not find corpus callosum find_cutting_plane:seed point not in structure! Searching neighborhood... recon-all -s blob2test exited with ERRORS at Tue Sep 14 13:18:14 CEST 2010
After this error I used the command: "recon-all -fill -cc-crs -pons-crs -lh-crs -rh-crs -noaseg -s blob2test" and got a bus error:
/Applications/freesurfer/subjects/blob2test/mri \n mri_fill -a ../scripts/ponscc.cut.log -Cv -pons-crs -lh-crs -rh-crs -xform transforms/talairach.lta wm.mgz filled.mgz \n logging cutting plane coordinates to ../scripts/ponscc.cut.log... using voxel position ( 0, 0, 0) as corpus callosum seed point INFO: Using transforms/talairach.lta and its offset for Talairach volume ... reading input volume...done. searching for cutting planes...voxel to talairach voxel transform 2.694 0.124 -0.276 -194.089; -0.311 2.259 -0.788 -3.124; 0.090 0.794 2.183 -266.051; 0.000 0.000 0.000 1.000; Bus error recon-all -s blob2test exited with ERRORS at Tue Sep 14 13:20:58 CEST 2010
My suspicion is that the way I have converted my WM volume is not correct and since the -fill uses the wm.mgz I do get error on this step. However I also tried to "align" my wm.mgz with brain.mgz using: "mri_convert mywm.mgz mynewwm.mgz --like brain.mgz". The "--apply_transform" option is not really clear to me and I do not get any good results if I apply that to mywhitematter.mgz. I also used tkregister2 to align the two volumes: my white matter with brain.mgz for example. But the wm.mgz file I've saved after registration to brain.mgz it could not be opened in tkmedit anymore so I decided not to use it as wm.mgz further.
I do want to use my white matter segmentation and continue with the steps in order to calculate with FS a pial surface, inflated surface, gyrification index, etc. for both hemispheres. I am also aware of the fact that using an adult atlas for baby scans is not correct so this might be also causing errors. One next step would probable be to make a baby brain atlas and include that instead of the */average/*.gca files.
What do you suggest as first starting point in solving the errors?
Thanks a million for your answer !!
Best regards,
Sabina