Martin,
As lgi is read like a thickness file, the lgi values in your tabular output replaced the value where you had thickness before. So mean lgi is in column 4 (note that average lgi values per parcell are comprised between 1 and 5 if your lgi computation is ok). Standard deviation for lgi is in column 5. The other columns should not change if you run mris_anatomical_stats without the -t option (i.e on the thickness).
About your last point, the main difference between gyrification index as previously computed (e.g. Zilles 1988) and local gyrification index is indeed the increased spatial resolution: you get one lgi value per cortical vertex (i.e. >100'000 lgi values over the hemisphere). Thus, the lgi per parcell from mris_anatomical_stats is the average lgi over all vertices included in each cortical parcel. Does it make sense?
Marie
On 9 mai 08, at 14:21, Martin Kavec wrote:
Thanks a lot help, Marie.
Indeed the -t switch worked.
Now, the lh.aparc.stats (thickness) and my generated lh.lgi.stats (lGI) differ in the columns 4,5,6, which correspond to GrayVol, AverageThk, and StdevThk. Could you please help me to understand these values?
Why is the GM volume different in the two files (column 4)?
Is the average lGI in the column 5?
My impression is that the lGI is a property of an area, e.i. of the
lateraloccipital cortex, so I would expect to get only a single value for the area.
Thanks a lot in advance,
Martin
On Friday 09 May 2008 10:53:13 Marie Schaer wrote:
Martin,
Try instead the option -t pial_lgi as follows:
mris_anatomical_stats -mgz -f lh.lgi.stats -b -a ../label/lh.aparc.annot -c ../label/ lgi.annot.ctab -t pial_lgi $SUBJECT lh
Otherwise it will read lh.pial_lgi as a surface file, whereas it is indeed a thickness / curv file
Have a nice day,
Marie
On 9 mai 08, at 10:29, Martin Kavec wrote:
Thanks for responce, Dough.
I've looked at recon-all as you suggested, and came up with the following commandline:
mris_anatomical_stats -mgz -f lh.lgi.stats -b -a ../label/lh.aparc.annot -c ../label/ lgi.annot.ctab $SUBJECT lh pial_lgi
This, however yields segfault:
INFO: assuming MGZ format for volumes. computing statistics for each annotation in ../label/lh.aparc.annot. reading volume $SUBJECT/mri/wm.mgz... reading input surface $SUBJECT/surf/lh.pial_lgi... Segmentation fault
I'm wondering whether this could be related to the fact that mris_anatomical_stats, calculates mean and stdev on the per-vertex basis over the defined label, while here we want it to calculate LGI, which is not calculated in this manner.
Thanks,
Martin
On Thursday 08 May 2008 18:32:28 Doug Greve wrote:
use mris_anatomical_stats. Look in recon-all to see how to invoke it. You can create another stats file (can't add it to the one that is already there).
doug
Martin Kavec wrote:
Hi,
is it possible to obtain LGI values of cortices, similarly as curvature indices in ?h.aparc.stats? Could this value possibly be included in the ?h.aparc.stats files?
Thanks in advance,
Martin
Freesurfer mailing list Freesurfer@nmr.mgh.harvard.edu https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer