[caret-users] mapping SPM2 activations volume to caret surface (activation in wrong place)

Donna Dierker donna at brainvis.wustl.edu
Sat Sep 5 07:56:40 CDT 2009


Soha,

This explains why you are having this problem.  You see, the SPM2 .hdr
format doesn't specify orientation information.  It depends on what your
analyze flip variable was set to in your SPM2 distribution, which at least
conceivably could change over time.  The data files themselves offer no
clues, unless you used something like a vitamin E capsule near the
subject's left temple, or something like that.  Caret doesn't right-left
flip the data, unless you open the .hdr via the File Open menu and check
the right-on-left checkbox as described below.

If you have lots of volumes you need to map, then this will get old fast. 
You'll want to look into converting these volumes to NIfTI.  There are a
number of command line utilities for doing this.  (I even think later
versions of caret_command might have such features.)

Donna

> the format is hdr/img not nii and I must use it (SPM2 data).. I will
> follow
> your instructions, thanks!
>
> Soha
>
> On Fri, Sep 4, 2009 at 6:46 PM, Donna Dierker
> <donna at brainvis.wustl.edu>wrote:
>
>> You didn't specify the format -- .nii or .hdr/.img.  Try to write your
>> volume as .nii and see how it maps.  If you must use .hdr/.img (e.g., it
>> really is SPM2 data), then use File: Open Data File: Volume Functional
>> file to open the volume.  Specify the data is right on left (checkbox on
>> file open dialog, I believe).  Then map loaded volume, rather than
>> volume
>> from disk.
>>
>> > Hi Donna:
>> >
>> > The is Caret version 5. I used the same version with another set of
>> data
>> > analyzed on SPM5 and no problem, it is mostly as you expected,
>> hemispheres
>> > flipped, can I flip them back on caret using the same spmT files?
>> >
>> > Thank you!
>> >
>> > Soha
>> > On Fri, Sep 4, 2009 at 3:21 PM, Donna Dierker
>> > <donna at brainvis.wustl.edu>wrote:
>> >
>> >> I'm not sure how old your Caret version is, but older versions of
>> Caret
>> >> couldn't read NIfTI files as .hdr/.img; you had to write them as .nii
>> to
>> >> get sensible results.  I suspect the left is getting mapped to right
>> and
>> >> vice versa, because your Caret version isn't reading the NIfTI volume
>> >> properly.
>> >>
>> >> If you're not using NIfTI, then fixing that is step 1.
>> >>
>> >> On 09/04/2009 01:28 PM, soha saleh wrote:
>> >> > Dear Caret users:
>> >> >
>> >> > I am trying to import fMRI data from a group study analyzed using
>> SPM2
>> >> > as a metric file so that I can view it on the PALS atlas. *However,
>> >> > the activations are shown in the wrong anatomical place!*
>> >> >
>> >> > This is what I do to map the volume to surface and to open the spec
>> >> file:
>> >> >
>> >> > Attributes\Map volumes to surface
>> >> > data mapping type : metric
>> >> > add volumes from disk : the spmT files Umt_Pmt_mask
>> >> > map to spec with atlas:
>> >> >                                  Output spec file: select
>> >> > Human.PALS_B12.BOTH.TEMPLATE-for-fMRI-MAPPING.73730.spec from this
>> >> > tutorial ( CARET_TUTORIAL_SEPT06/MAPPING_PROCEDURES)
>> >> >                                  Mapping Atlas: SPM2
>> >> >                                  Atlas:
>> >> > Human.PALS_B12.BOTH.TEMPLATE-for-fMRI-MAPPING.73730.spec
>> >> > Mapping algorithm: Metric enclosing voxel
>> >> >
>> >> > Now load spec file
>> >> > file\open spec file :
>> >> > Human.PALS_B12.BOTH.TEMPLATE-for-fMRI-MAPPING.73730.spec
>> >> > SPACE: SPM2
>> >> > category: atlas
>> >> > species: human
>> >> > subject: PALS-B12
>> >> >
>> >> > select (Fiducial) and (Inflated) coordinates, and the metric file
>> >> > created after I did map volume to surface. I convert metric to RGB,
>> I
>> >> > overlay the RGB data, and I have wrong activations.
>> >> >
>> >> > I attached a ppt. file that has 2 slides each showing the same spmT
>> >> > file on either Caret or SPM2 result figure, both at threshold T=3.
>> >> >
>> >> > Between, I did the same procedure with data analyzed using SPM5 and
>> >> > had no problem.
>> >> >
>> >> > I hope anybody can help, thanks in advance!
>> >> >
>> >> > Soha Saleh
>> >> > phD candidate
>> >> > Department of Biomedical Engineering
>> >> > New Jersey Institute of Technology
>> >> > shs25 at njit.edu <mailto:shs25 at njit.edu>
>> >> > 1 (201) 951 0029
>> >> >
>> >> >
>> >> >
>> >> >
>> >> >
>> >> >
>> ------------------------------------------------------------------------
>> >> >
>> >> > _______________________________________________
>> >> > caret-users mailing list
>> >> > caret-users at brainvis.wustl.edu
>> >> > http://brainvis.wustl.edu/mailman/listinfo/caret-users
>> >>
>> >> _______________________________________________
>> >> caret-users mailing list
>> >> caret-users at brainvis.wustl.edu
>> >> http://brainvis.wustl.edu/mailman/listinfo/caret-users
>> >>
>> > _______________________________________________
>> > caret-users mailing list
>> > caret-users at brainvis.wustl.edu
>> > http://brainvis.wustl.edu/mailman/listinfo/caret-users
>> >
>>
>> _______________________________________________
>> caret-users mailing list
>> caret-users at brainvis.wustl.edu
>> http://brainvis.wustl.edu/mailman/listinfo/caret-users
>>
> _______________________________________________
> caret-users mailing list
> caret-users at brainvis.wustl.edu
> http://brainvis.wustl.edu/mailman/listinfo/caret-users
>




More information about the caret-users mailing list