Dear AFNI experts,
I have a problem with the mapping between a surface and a volume in SUMA and AFNI; specifically, coordinate transformations between surface and volume seems off. I just started using surfaces, so maybe it's a simple beginner's mistake; nevertheless it would be great if someone could help to understand what's going wrong.
In short, I made a left hemisphere surface model with caret using a T1 scan, which was first aligned to talairach space in AFNI, oriented in LPI, converted to .nii, and cut in caret to preserve only the left hemisphere. This worked well, and in caret I can overlay a functional brik (in tlrc space) that shows activations in the same locations as the volume view in AFNI.
Using the @SUMA_Make_Spec_Caret script I made a spec file for suma; this worked after I made the script a bit more lenient towards caret5's naming conventions. Using this spec file, the caret surface models can be viewed in suma as expected. However, after 'talking' between afni and suma has been enabled, right-clicking on the surface puts the cross-hair in the corresponding volume way outside the brain. For example, clicking at an arbitrary position on the surface shows in the terminal window:
FaceSet 109709, Closest Node 54729
Nodes forming closest FaceSet:
54729, 54724, 53936
Coordinates of Nodes forming closest FaceSet:
84.795258, 54.434666, 9.006767
[...]
while in the caret .coord file, the corresponding volume coordinate for node 54729 ("N1") is:
54729 -36.047188 8.113117 55.786102
The volume brik is well aligned (origin at AC), so something seems to be going wrong in the coordinate transformation. The coordinate problem does not seem to be just an offset problem, as picking another point ("N2") on the surface and taking the difference between N1 and N2 both in surface and volume coordinates does not yield the same difference vector.
Finally, suma complaints that
"Idetincal values for AC and AC_WholeVolume.
Check your params file if not using Talairach-ed surfaces."
however this should be fine because the volume is in tlrc space.
Below are suma's output, contents of the spec file, and the params file that the spec file refers to.
Any ideas?
Thanks
Nick
*** SUMA output ***
suma -spec sr906_lh.spec -sv anat_avg_lpi_rs-LEFT+tlrc.HEAD
vvvvvvvvvvvvvvvvvvvvvvvvvvvvSurface #1/3(Local Domain Parent), loading ...
Warning SUMA_Read_SureFit_Param: All values for AC are 0.0.
Warning SUMA_Read_SureFit_Param: All values for AC_WholeVolume are 0.0.
oo Warning SUMA_Read_SureFit_Param (SUMA_Surface_IO.c:1321):
Idetincal values for AC and AC_WholeVolume.
Check your params file if not using Talairach-ed surfaces.
vvvvvvvvvvvvvvvvvvvvvvvvvvvvSurface #2/3 (mappable via Local Domain Parent), loading ...
Warning SUMA_Read_SureFit_Param: All values for AC are 0.0.
Warning SUMA_Read_SureFit_Param: All values for AC_WholeVolume are 0.0.
oo Warning SUMA_Read_SureFit_Param (SUMA_Surface_IO.c:1321):
Idetincal values for AC and AC_WholeVolume.
Check your params file if not using Talairach-ed surfaces.
vvvvvvvvvvvvvvvvvvvvvvvvvvvvSurface #3/3 (mappable via Local Domain Parent), loading ...
Warning SUMA_Read_SureFit_Param: All values for AC are 0.0.
Warning SUMA_Read_SureFit_Param: All values for AC_WholeVolume are 0.0.
oo Warning SUMA_Read_SureFit_Param (SUMA_Surface_IO.c:1321):
Idetincal values for AC and AC_WholeVolume.
Check your params file if not using Talairach-ed surfaces.
SUMA_niml_call: Contacting ...
SUMA_niml_call: Trying local connection...
++ NIML connection opened from 127.0.0.1
SUMA_Engine: Sending surface Human.sr906.L.Fiducial.57654.coord__Human.sr906.L.Fiducial.57654.topo (NodeList, FaceSetList, NodeNormList)...
+++ NOTICE: SUMA_ixyz: replaced 57654 nodes
for surface Human.sr906.L. (#0),
session /Users/psyuser/Documents/organized/105_observe_execute_actions2/caret/t1/
I/O time = 59 ms, Processing = 2 ms
+++ NOTICE: 57654 normals attached
to surface Human.sr906.L. (#0),
session /Users/psyuser/Documents/organized/105_observe_execute_actions2/caret/t1/
I/O time = 6 ms, Processing = 1 ms
SUMA_Engine: Sending switch underlay command to (SWITCH_UNDERLAY A.anat_avg_lpi_rs-LEFT)...
*** sr906_lh.spec ***
# delimits comments
# Creation information:
# user : psyuser
# date : Sun 17 May 2009 16:50:27 BST
# machine : pmac2774.bangor.ac.uk
# pwd : /Users/psyuser/Documents/organized/105_observe_execute_actions2/caret/t1
# command : @SUMA_Make_Spec_Caret -debug 1 -sid sr906
# define the group
Group = sr906
# define various States
StateDef = Fiducial
StateDef = Raw
StateDef = VeryInflated
StateDef = Inflated
NewSurface
SurfaceFormat = ASCII
SurfaceType = SureFit
SureFitCoord = Human.sr906.L.Fiducial.57654.coord
SureFitTopo = Human.sr906.L.Fiducial.57654.topo
SureFitVolParam = Human.sr906.L.params_file_22.params
LocalDomainParent = SAME
SurfaceState = Fiducial
EmbedDimension = 3
NewSurface
SurfaceFormat = ASCII
SurfaceType = SureFit
SureFitCoord = Human.sr906.L.VeryInflated.57654.coord
SureFitTopo = Human.sr906.L.Fiducial.57654.topo
SureFitVolParam = Human.sr906.L.params_file_22.params
LocalDomainParent = Human.sr906.L.Fiducial.57654.coord
SurfaceState = VeryInflated
EmbedDimension = 3
NewSurface
SurfaceFormat = ASCII
SurfaceType = SureFit
SureFitCoord = Human.sr906.L.Inflated.57654.coord
SureFitTopo = Human.sr906.L.Fiducial.57654.topo
SureFitVolParam = Human.sr906.L.params_file_22.params
LocalDomainParent = Human.sr906.L.Fiducial.57654.coord
SurfaceState = Inflated
EmbedDimension = 3
*** Human.sr906.L.params_file_22.params ***
BeginHeader
Caret-Version 5.61
comment
date Sun May 17 16:55:12 2009
encoding ASCII
pubmed_id
EndHeader
ACx=
ACx_WholeVolume=
ACy=
ACy_WholeVolume=
ACz=
ACz_WholeVolume=
CGMpeak=39
CropMaxX=88
CropMaxY=190
CropMaxZ=230
CropMinX=10
CropMinY=0
CropMinZ=95
OldPadNegX=
OldPadNegY=
OldPadNegZ=
OldPadPosX=
OldPadPosY=
OldPadPosZ=
WMThreshSet=0
WMpeak=83
WholeXdim=
WholeYdim=
WholeZdim=
Xmin=10
Ymin=0
Zmin=95
comment=
cropped=yes
padded=
resolution=
xdim=
ydim=
zdim=
