Re: [HCP-Users] Surface files for probtrackx

Thu, 17 Jul 2014 04:32:53 -0700

Mesh convention should be caret.

Matt.

From: Saad Jbabdi <[email protected]>
Date: Thursday, July 17, 2014 at 6:08 AM
To: Patricia Hong Phi Tran <[email protected]>
Cc: "[email protected]" <[email protected]>
Subject: Re: [HCP-Users] Surface files for probtrackx

Hi - the surface reference should, in your case, be a MNI space volume file. 

Cheers
Saad

On 17 Jul 2014, at 05:31, Patricia Hong Phi Tran <[email protected]> wrote:

Hi,

Thank you very much for this, we got a little bit further but are still having some troubles.
First off, we ran the following syntax in accordance with your instructions and were able to successfully visualise the ROIs using the workbench

>> wb_command -gifti-label-to-roi MNINonLinear/fsaverage_LR32k/100307.L.aparc.32k_fs_LR.label.gii L.frontalpole.func.gii -name L_frontalpole
>> surf2surf -i /MNINonLinear/fsaverage_LR32k/100307.L.white.32k_fs_LR.surf.gii -o L.frontalpole.gii --values=L.frontalpole.func.gii

>> wb_command -gifti-label-to-roi MNINonLinear/fsaverage_LR32k/100307.L.aparc.32k_fs_LR.label.gii L.temporalpole.func.gii -name L_temporalpole
>> surf2surf -i /MNINonLinear/fsaverage_LR32k/100307.L.white.32k_fs_LR.surf.gii -o L.temporalpole.gii --values=L.temporalpole.func.gii

Following this we tried to submit these ROIs (L.frontalpole.gii and L.temporalpole.gii) to probtrackx2:

>> probtrackx2 --network -x 100307/T1w/Diffusion.probtrackX/surfaceconnectome/masks.txt -V 1 -l --onewaycondition --omatrix1 -c 0.2 -S 2000 --steplength=0.5 -P 5000 --fibthresh=0.01 --distthresh=0.0 --sampvox=0.0 —xfm=/100307/MNINonLinear/xfms/standard2acpc_dc.nii.gz —invxfm=/100307/MNINonLinear/xfms/acpc_dc2standard.nii.gz --forcedir --opd -s 100307/T1w/Diffusion.bedpostX/merged -m /100307/T1w/Diffusion.bedpostX/nodif_brain_mask —dir=/100307/T1w/Diffusion.probtrackX/surfaceconnectome 

However, we’re unsure about which image to provide as a “Surface Reference Image” for the Freesurfer mesh convention. We tried a host of different files in 100307’s directory and each time we received no output from probtrackx2. Finally, we tried ignoring the GUIs request for a surface reference image and we (unsurprisingly) got a warning about this but then we received some plausible output (see attached).

Of course, if we don’t select “seed space is not diffusion” we don’t seem to need to provide the reference image at all but since we used the .label.gii and .surf.gii files from 100307’s MNINonLinear directory we figured we ought to supply the relevant transforms.

Thanks,

- Patricia


On 16 July 2014 19:27, Saad Jbabdi <[email protected]> wrote:
Hi

In the surf2surf command, the input should be a surf.gii file and the output a gii file (without the “surf”), which will combine information from both the surf.gii file (the actual surface)  and the func.gii file (which contains your ROI).

You can visualise your ROI in either workbench or fslview. 

In workbench, simply load the surf.gii and the func.gii.

In FSL, you need to first project the output of surf2surf onto voxels:
surf2volume L.frontalpole.gii SomeVolumeInMNISpace.nii.gz output.nii.gz caret
fslview output.nii.gz

Cheers
Saad






On 16 Jul 2014, at 05:37, Patricia Hong Phi Tran <[email protected]> wrote:

Hi,

We are trying to run surface-to-surface tractography using probtrackx and generate whole-brain connectomes with the HCP data but have run into a number of problems. 
We attempted to follow the wiki but label2surf only works with Freesurfer label files, not the GIFTI ones included. 

Therefore;
1) We used wb_command -gifti-label-to-roi with the L.aparc.32k_fs_LR.label GIFTI file to extract the frontal and temporal pole into respective GIFTI labels (e.g. lh.frontalpole.func.gii and lh.temporalpole.func.gii). We are not sure that this worked as we cannot visualise the files in either the workbench or FSL due to incompatibility issues. 
wb_command -gifti-label-to-roi 100307/MNINonLinear/fsaverage_LR32k/100307.L.aparc.32k_fs_LR.label.gii  /100307/L.frontalpole.func.gii -name L_frontalpole

2) Then we attempted to generate surface/label files from these individual label files using surf2surf (e.g. lh.frontalpole.func.gii --> lh.frontalpole.surf.gii)
surf2surf -i L.frontalpole.func.gii -o L.frontalpole.surf.gii --values=L.frontalpole.func.gii 

3) Fed in both the frontal and temporal pole .surf.gii files to probtrackx2 using the "multiple masks" option in the GUI, selecting Matrix1 as an option with otherwise standard settings.

4) It ran for ~66,000 seconds and then 'finished' with the following error message (attached) and the matrix output consisted of 0s, indicating that no tracts appeared to have been run at all.

It ultimately seems to be a problem with the GIFTI label extraction and conversion but we are not sure what the fix is.

Thanks in advance, 

- Patricia

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

<Screen Shot 2014-07-16 at 9.31.51 AM.jpg>

--
Saad Jbabdi
University of Oxford, FMRIB Centre

JR Hospital, Headington, OX3 9DU, UK
(+44)1865-222466  (fax 717)


<fdt_paths.nii.gz>

--
Saad Jbabdi
University of Oxford, FMRIB Centre

JR Hospital, Headington, OX3 9DU, UK
(+44)1865-222466  (fax 717)

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

 

The materials in this message are private and may contain Protected Healthcare Information or other information of a sensitive nature. If you are not the intended recipient, be advised that any unauthorized use, disclosure, copying or the taking of any action in reliance on the contents of this information is strictly prohibited. If you have received this email in error, please immediately notify the sender via telephone or return mail.

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

Reply via email to