Print

Print
A couple of comments:
(1) The smoothness in SPM can be found in SPM.mat file --. SPM.xVol.FWHM.
(2) spm_estimate_smoothness uses the actual residual images. The ResMS is the mean of the squares of the residuals.
(3) spm_estimate returns a value in voxels
(4) For AFNI, you should take the squareroot of the ResMS image first to get something more similar to the individual residual images.

Best Regards, Donald McLaren
=================
D.G. McLaren, Ph.D.
Research Fellow, Department of Neurology, Massachusetts General Hospital and
Harvard Medical School
Postdoctoral Research Fellow, GRECC, Bedford VA
Website: http://www.martinos.org/~mclaren
Office: (773) 406-2464
=====================
This e-mail contains CONFIDENTIAL INFORMATION which may contain PROTECTED
HEALTHCARE INFORMATION and may also be LEGALLY PRIVILEGED and which is
intended only for the use of the individual or entity named above. If the
reader of the e-mail is not the intended recipient or the employee or agent
responsible for delivering it to the intended recipient, you are hereby
notified that you are in possession of confidential and privileged
information. Any unauthorized use, disclosure, copying or the taking of any
action in reliance on the contents of this information is strictly
prohibited and may be unlawful. If you have received this e-mail
unintentionally, please immediately notify the sender via telephone at (773)
406-2464 or email.

On Fri, Oct 10, 2014 at 1:43 PM, John Flournoy <[log in to unmask]> wrote:
Hi All,

I've been scouring the Internet and the lit for how to properly estimate smoothness for finding empirical cluster/p-value combos for FDR correction. Many sources indicate one should average the estimated FWHM for first-level (that is, individual level) model residuals. Hopefully I'm correct that the residuals are in ResMS.nii. This seems pretty straightforward using AFNI's 3dFWHMx, but I'd like to automate the process using SPM and MATLAB, so I've been looking into the `spm_est_smoothness`. Below are the return values for the SPM function and for the AFNI program, which are wildly different (notice that spm_est_smoothness FWHM is of the order 10^3:

SPM:

>> [FWHM,VRpv,R] = spm_est_smoothness('ResMS.nii','mask.nii')
Spatial non-sphericity (over scans)     :                        ...done

FWHM =

   1.0e+03 *

    1.0528    1.1365    1.1931


VRpv =

      fname: 'RPV.nii'
        dim: [53 63 52]
         dt: [64 0]
        mat: [4x4 double]
      pinfo: [3x1 double]
    descrip: 'spm_spm: resels per voxel'
          n: [1 1]
    private: [1x1 nifti]


R =

    2.0000    0.1286    0.0048    0.0000

AFNI:

$ 3dFWHMx -mask mask.nii ResMS.nii
++ 3dFWHMx: AFNI version=AFNI_2011_12_21_1014 (Jun 16 2014) [64-bit]
++ Authored by: The Bob
++ Number of voxels in mask = 60139
** AFNI converts NIFTI_datatype=64 (FLOAT64) in file ResMS.nii to FLOAT32
     Warnings of this type will be muted for this session.
     Set AFNI_NIFTI_TYPE_WARN to YES to see them all, NO to see none.
 8.85452  9.52384  8.40267

Does anyone know what's going on here? Thanks much for any insight!

~j