It estimates the FWHM from the data (glmdir/fwhm.dat). It does this by computing the pearson correation coefficient of the residual error between adjacent vertices. This is then averaged over all vertices to compute a mean cor coef. A transform is then applied to compute the FWHM. This is consistent in that if you synthesize white gaussian noise on the surface, then smooth it with a given FWHM (mri_surf2surf or mris_fwhm), then use that data to compute the FWHM, the applied and estimated FWHM will be equal (or close).
If you look in a given file produced by the simulator, eg, $FREESURFER_HOME/average/mult-comp-cor/fsaverage/lh/cortex/fwhm10/abs/th20/mc-z.cdf the two relevant columns will be MaxClustBin and MaxClustCDF, eg, # BinNo MaxClustBin MaxClustPDF MaxClustCDF GRFCDF MaxSigBin MaxSigPDF MaxSigCDF 33 181.572128 0.010100 0.092500 0.132425 14.427808 0.004600 0.593700 Means that a cluster of size 181.57 has a probability of 0.0925 under the null when the FWHM=10 and the threshold is p<.01 (th20) On 05/06/2016 03:59 PM, Ajay Kurani wrote: > Hi Doug, > I wanted to figure out what is the minimum cluster extent for a > given monte carlo simulation (ie when you choose p=0.5/T=1.3). I saw > the following tutorial : > https://surfer.nmr.mgh.harvard.edu/fswiki/BuildYourOwnMonteCarlo and > so I wanted to know if freesurfer estimates the smoothness for each > subject's cortical thickness for a hemisphere and then chooses the > corresponding fwhm folder or if by choosing lh 15mm in qdec the > clusterwise values are taken from the 15mm folder? Is the smoothness > based on the smoothing kernel used or the inherent smoothness of that > particular measure (ie measuring the smoothness of the input > thicknesss image)? Is there a way to see what cluster minimum is used? > > Thanks, > Ajay > > > On Thu, May 5, 2016 at 3:49 PM, Ajay Kurani <dr.ajay.kur...@gmail.com > <mailto:dr.ajay.kur...@gmail.com>> wrote: > > Hi Doug, > Thanks for the clarification. So in the case of cortical > thickness, qcache, mris_fwhm or mri_surf2surf would all do the > same thing and so I should be getting similar results if > everything is entered in the same fashion. This would be the > approriate choice compared to mri_smooth. > > For mri_surf2surf I used the following command for smoothing LGI > and cortical thickness and converting to .gii files. > > mris_surf2surf --prune --s fsaverage --hemi rh --fwhm 15 --sval > rh.thickness.fsaverage.mgh --tval rh.thickness.fwhm15.mgz --cortex > mris_convert -c rh.thickness.fwhm15.mgz > $FREESURFER_HOME/subjects/fsaverage/surf/rh.white > rh.thickness.fwhm15.gii > > > 1) For cortical thickness does it make sense to use the --cortex > option or should I specify a mask of some type (if so which) in > mris_surf2surf? > > 2) For converting files to .gii should I be using rh.white as the > option or should it be rh.pial? > > > Best, > Ajay > > > > On Thu, May 5, 2016 at 2:53 PM, Ajay Kurani > <dr.ajay.kur...@gmail.com <mailto:dr.ajay.kur...@gmail.com>> wrote: > > Hi Doug, > Thanks for the quick reply. > > Is there a difference from qcache/mris_fwhm with mris_smooth > and mri_surf2surf -fwhm ? If so, which is recommended for > cortical thickness analysis? > > Thanks, > Ajay > > On Wed, May 4, 2016 at 11:18 PM, Ajay Kurani > <dr.ajay.kur...@gmail.com <mailto:dr.ajay.kur...@gmail.com>> > wrote: > > Hi Freesurfer Experts, > Just as a followup through my reading i've come across > posts which use qcache, mris_fwhm, mri_surf2surf or > mris_smooth for smoothing. For my cortical thickness > analysis I would like to smooth all of my > rh/lh.thickness.fsaverage.mgh files for each subject in > order to run a group analysis. After finding regions of > difference, I would then like to use the ROI to extract > each individual's mean thickness in the ROI in order to > run a correlation with other measures. Based on this, I > assume it would make sense to use smoothed data to > identify the ROI and then use unsmoothed data for > extracting actual thickness measures (does > lh.thickness.fsaverage contain the original thickness or > warped thickness values). > > I am unsure which smoothing is the most accurate or > preferred. In using qcache the smoothness of the images > do not seem to reach the filter level (based on the > earlier email) so I am not sure if there is a freesurfer > tool to check the smoothness level or if the qcache > smoothness levels make sense for cortical thickness. > > Thanks, > Ajay > > On Wed, May 4, 2016 at 5:41 PM, Ajay Kurani > <dr.ajay.kur...@gmail.com > <mailto:dr.ajay.kur...@gmail.com>> wrote: > > Hi Freesurfer Experts, > I am trying to understand the difference between > qcache option and mris_fwhm and which is appropriate > for a cortical thickness analysis. I processed my > files with qcache and have > lh.thickness.fsaverage.fwhm15.gii (converted) files. I > used an afni tool SurfFWHM to estimate the smoothness > of a subject at when looking at the fwhm0 image it > iwas 5.5 and for 10, 15 and 20mm it was approximately > 9.3-9.9 smoothness level. > > I also used mris_fwhm --hemi lh --s fsaverage > --smooth-only --i lh.thickness.fsaverage.mgz --fwhm 15 > --cortex --o test_15.gii and when using SurfFWHM on > the smae subject the smoothness was estimated at 11.25. > > > 1) I am not sure if the qcache or the mris_fwhm file > is more appropriate to use for a cortical thickness > analysis. > > 2) For qdec if I select the 15mm option does it > assume the smoothness is 15mm when calculating monte > carlo corrections? Would there be a different way to > estimate this since my smoothness at 15mm is closer to > 10mm? > > Thanks, > Ajay > > > > > > > > _______________________________________________ > Freesurfer mailing list > Freesurfer@nmr.mgh.harvard.edu > https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer -- Douglas N. Greve, Ph.D. MGH-NMR Center gr...@nmr.mgh.harvard.edu Phone Number: 617-724-2358 Fax: 617-726-7422 Bugs: surfer.nmr.mgh.harvard.edu/fswiki/BugReporting FileDrop: https://gate.nmr.mgh.harvard.edu/filedrop2 www.nmr.mgh.harvard.edu/facility/filedrop/index.html Outgoing: ftp://surfer.nmr.mgh.harvard.edu/transfer/outgoing/flat/greve/ _______________________________________________ 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.
