When you run mri_glmfit, it computes an estimate of the FWHM from the residuals. In this case, the FWHM=46 eventhough you smoothed it by only 15. This can occur for a couple of reasons. One is that one or more subjects have bad data, another is that the data are inherently very smooth. Try running it without any applied smoothing and see what FWHM you get (glmdir/fwhm.dat). You can also check your input data to see if any subjects are way off from others.

doug

On 02/09/2016 08:40 AM, Domenico Zacà wrote:

Dear FS team,

I am running a surface based analysis on GWR on a group of 203 subjects. I generated left and right GWR maps, smoothed with a Gaussian kernel of FWHM 15 mm. However when I try to run the multiple comparisons correction analysis (mri_glmfit-sim --glmdir lh.w-g.pct_aesint.15.glmdir --cache 2 pos --cwp 0.05 --2spaces) I get the following error message:

fwhm = 46.240061 ERROR: cannot find /opt/freesurfer/freesurfer-stable5/average/mult-comp-cor/fsaverage/lh/cortex/fwhm46/pos/th20/mc-z.csd

Any idea how to get around this?

Thanks

Domenico Zacà, PhD

MR Lab Research Fellow

Center for Mind/Brain Sciences

University of Trento, Italy

Via delle Regole, 101

38100 Mattarello (TN)

Italy

Telephone: +39-0461-28 3069 tel:%2B39-0461-28%203069 Fax: +39-0461-28-3066 tel:%2B39-0461-28-3066

Email: domenico.zaca@unitn.it mailto:domenico.zaca@unitn.it

MRI Methods Group: http://r.unitn.it/en/cimec/mri

---

Freesurfer mailing list Freesurfer@nmr.mgh.harvard.edu https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer