Hi Lars,

this test is too sensitive in FS5.3 . We have changed that a while ago. The mri_diff command should be
mri_diff --notallow-pix --notallow-geo ....
if that still shows a difference, something is different in the image resolution and that could be a problem. Only differences in the transformation matrices are no problem.

Best, Martin


On 04/22/2016 08:52 AM, Lars M. Rimol wrote:
Hi,

I'm running longitudinal recon-all (freesurfer-Linux-centos6_x86_64-stable-pub-v5.3.0) on Ubuntu 14.04.4, with three timepoints per subject. I get this warning about geometry (please see below). All three images have been converted from dicom to .mgz using mri_convert. From a previous question on the list from last year, I get the impression this warning should not appear if all images have been transformed the same way. I've pasted mri_info on both images at the bottom of the e-mail. The transformation matrices differ from timepoint to timepoint and I suspect this might be the reason for the warning. Is this a cause for concern?

Thank you!



#@# Longitudinal Base Subject Creation fr. 22. april 09:07:47 +0200 2016
/media/lmr2/subjects/TBI/artifakt_qc/KT5065_base

 mri_diff --notallow-pix /media/lmr2/subjects/TBI/artifakt_qc/Kt5065_2/mri/rawavg.mgz /media/lmr2/subjects/TBI/artifakt_qc/KT5065_1/mri/rawavg.mgz

Volumes differ in geometry row=1 col=4 diff=1.606422 (1.60642)

 mri_diff --notallow-pix /media/lmr2/subjects/TBI/artifakt_qc/kt5065_reskann_2/mri/rawavg.mgz /media/lmr2/subjects/TBI/artifakt_qc/KT5065_1/mri/rawavg.mgz

Volumes differ in geometry row=1 col=4 diff=1.713524 (1.71352)

*******************************************************************************
WARNING: Image geometries differ across time, maybe due to aquisition changes?
         This can potentially bias a longitudinal study! Will continue in 10s.
*******************************************************************************




---------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
Comparison using mri_info

[09:27 lmr@dmed4870] > mri_info KT5065_1/mri/orig/001.mgz
Volume information for KT5065_1/mri/orig/001.mgz
          type: MGH
    dimensions: 256 x 256 x 176
   voxel sizes: 1.0000, 1.0000, 1.0000
          type: SHORT (4)
           fov: 256.000
           dof: 0
        xstart: -128.0, xend: 128.0
        ystart: -128.0, yend: 128.0
        zstart: -88.0, zend: 88.0
            TR: 2300.00 msec, TE: 4.21 msec, TI: 996.00 msec, flip angle: 9.00 degrees
       nframes: 1
       PhEncDir: UNKNOWN
ras xform present
    xform info: x_r =  -0.0000, y_r =  -0.0000, z_r =  -1.0000, c_r =    -0.5000
              : x_a =  -1.0000, y_a =  -0.0000, z_a =  -0.0000, c_a =     0.0000
              : x_s =   0.0000, y_s =  -1.0000, z_s =   0.0000, c_s =     0.0000

talairach xfm :
Orientation   : PIL
Primary Slice Direction: sagittal

voxel to ras transform:
               -0.0000  -0.0000  -1.0000    87.5000
               -1.0000  -0.0000  -0.0000   128.0000
                0.0000  -1.0000   0.0000   128.0000
                0.0000   0.0000   0.0000     1.0000

voxel-to-ras determinant -1

ras to voxel transform:
               -0.0000  -1.0000  -0.0000   128.0000
               -0.0000  -0.0000  -1.0000   128.0000
               -1.0000   0.0000   0.0000    87.5000
                0.0000   0.0000   0.0000     1.0000


[09:27 lmr@dmed4870] > mri_info Kt5065_2/mri/orig/001.mgz
Volume information for Kt5065_2/mri/orig/001.mgz
          type: MGH
    dimensions: 256 x 256 x 176
   voxel sizes: 1.0000, 1.0000, 1.0000
          type: SHORT (4)
           fov: 256.000
           dof: 0
        xstart: -128.0, xend: 128.0
        ystart: -128.0, yend: 128.0
        zstart: -88.0, zend: 88.0
            TR: 2300.00 msec, TE: 4.21 msec, TI: 996.00 msec, flip angle: 9.00 degrees
       nframes: 1
       PhEncDir: UNKNOWN
ras xform present
    xform info: x_r =  -0.0000, y_r =  -0.0000, z_r =  -1.0000, c_r =    -2.1064
              : x_a =  -1.0000, y_a =  -0.0000, z_a =  -0.0000, c_a =    -3.2129
              : x_s =   0.0000, y_s =  -1.0000, z_s =   0.0000, c_s =    -5.6225

talairach xfm :
Orientation   : PIL
Primary Slice Direction: sagittal

voxel to ras transform:
               -0.0000  -0.0000  -1.0000    85.8936
               -1.0000  -0.0000  -0.0000   124.7871
                0.0000  -1.0000   0.0000   122.3775
                0.0000   0.0000   0.0000     1.0000

voxel-to-ras determinant -1

ras to voxel transform:
               -0.0000  -1.0000  -0.0000   124.7871
               -0.0000  -0.0000  -1.0000   122.3775
               -1.0000   0.0000   0.0000    85.8936
                0.0000   0.0000   0.0000     1.0000


[09:28 lmr@dmed4870] > mri_info kt5065_3/mri/orig/001.mgz
Volume information for kt5065_3/mri/orig/001.mgz
          type: MGH
    dimensions: 256 x 256 x 176
   voxel sizes: 1.0000, 1.0000, 1.0000
          type: SHORT (4)
           fov: 256.000
           dof: 0
        xstart: -128.0, xend: 128.0
        ystart: -128.0, yend: 128.0
        zstart: -88.0, zend: 88.0
            TR: 2300.00 msec, TE: 4.21 msec, TI: 996.00 msec, flip angle: 9.00 degrees
       nframes: 1
       PhEncDir: UNKNOWN
ras xform present
    xform info: x_r =  -0.0000, y_r =  -0.0000, z_r =  -1.0000, c_r =     1.2135
              : x_a =  -1.0000, y_a =  -0.0000, z_a =  -0.0000, c_a =     8.4827
              : x_s =   0.0000, y_s =  -1.0000, z_s =   0.0000, c_s =   -11.9426

talairach xfm :
Orientation   : PIL
Primary Slice Direction: sagittal

voxel to ras transform:
               -0.0000  -0.0000  -1.0000    89.2135
               -1.0000  -0.0000  -0.0000   136.4827
                0.0000  -1.0000   0.0000   116.0574
                0.0000   0.0000   0.0000     1.0000

voxel-to-ras determinant -1

ras to voxel transform:
               -0.0000  -1.0000  -0.0000   136.4827
               -0.0000  -0.0000  -1.0000   116.0574
               -1.0000   0.0000   0.0000    89.2135
                0.0000   0.0000   0.0000     1.0000






sincerely yours,

Lars M. Rimol, PhD
Senior researcher,
Norwegian Advisory Unit for functional MRI
Department of Radiology,
St. Olav's University hospital,
7006 Trondheim,
Norway





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

-- 
Martin Reuter, PhD
Assistant Professor of Radiology, Harvard Medical School
Assistant Professor of Neurology, Harvard Medical School
A.A.Martinos Center for Biomedical Imaging
Massachusetts General Hospital
Research Affiliate, CSAIL, MIT
Phone: +1-617-724-5652
Web  : http://reuter.mit.edu