For the list's benefit.

Tim

---------- Forwarded message ----------
From: Maarten Vaessen <[email protected]>
Date: Thu, Nov 3, 2016 at 3:45 AM
Subject: Re: [HCP-Users] -probtrackx-dot-convert volume label problem
To: Timothy Coalson <[email protected]>


The problem was that the coords_for_fdt_matrix3 file (from probtrackx2) has
5 colums (x,y,z,something,something), while the -probtrackx-dot-convert expects
just the x,y,z in a file.
It works now!

Thanks,

-Maarten

On Wed, Nov 2, 2016 at 9:43 PM, Timothy Coalson <[email protected]> wrote:

> As a guess, did you use "-crop" when you used -cifti-separate to get the
> amygdala ROI?  If so, that changes the volume space (specifically the FOV),
> so indices don't correspond with the normal 2mm volume space.  This would
> mean that you now need to make a label volume in this new FOV space.
> Alternatively, you could get the ROI without using -crop, rerun probtrackx,
> and then the standard Atlas_ROIs file will work.
>
> The <label-vol> must have the same dimensions and origin/spacing as the
> roi file you used in probtrackx, otherwise the indices won't match.  I
> would not call an amygdala-only volume "Atlas_ROIs", that would be
> misleading, and you should double check your amygdala ROI file for its
> value, as it probably uses 1 rather than 18, and fix the labels text file
> based on that, but the -volume-label-import command otherwise looks like it
> would produce something that should work for your current probtrackx
> results.
>
> Tim
>
>
> On Wed, Nov 2, 2016 at 7:20 AM, Maarten Vaessen <[email protected]>
> wrote:
>
>> Hello,
>>
>> I'm trying to convert a .dot from probtrackx2 with -matrix3 option to a
>> dconn.nii. The -lrtarget was a list containing all volumes and surfaces
>> conforming to the HCP grayordinate file. The -target3 is the L (or R)
>> amygdala from Atlas_ROI.2.nii.gz. The output from probtrackx2 looks
>> sensible.
>>
>> I use the following command and get this error:
>> wb_command -probtrackx-dot-convert fdt_matrix3.dot fdt_matrix3.dconn.nii
>> -col-cifti 91282_Greyordinates.dscalar.nii COLUMN -row-voxels
>> coords_for_fdt_matrix3 amyL_label_from_Atlas.nii.gz
>>
>> ERROR: voxel index in list file did not match a structure: 0, 1, 56
>>
>>
>> Now, I am not sure what to use as input for -probtrackx-dot-convert
>> -row-voxels:
>>
>> [-row-voxels] - the output mapping along a row will be voxels
>>
>> <voxel-list-file> - a text file containing IJK indices for the voxels
>>  used
>> --> this should be the coords_for_fdt_matrix3 right?
>>
>> <label-vol> - a label volume with the dimensions and sform used, with
>>             structure labels
>> --> here I am a bit lost. I tried several options, the probably most
>> sensible being:
>> wb_command -volume-label-import subj/MNINonLinear/ROIs/Atlas_ROIs.2.nii
>> amyL_label.txt amyL_label_from_Atlas.nii.gz -discard-others
>>
>> amyL_label.txt =
>> AMYGDALA_LEFT
>> 18 103 255 255 255
>>
>> What would be the correct input to get this to work?
>>
>> Thx,
>>
>> -Maarten
>>
>> _______________________________________________
>> HCP-Users mailing list
>> [email protected]
>> http://lists.humanconnectome.org/mailman/listinfo/hcp-users
>>
>
>

_______________________________________________
HCP-Users mailing list
[email protected]
http://lists.humanconnectome.org/mailman/listinfo/hcp-users

Reply via email to