 |
 |
Thanks. I used mri_annotation2label to create the labels and use lh.lateralorbitofrontal.label and lh.superiortemporal.label to track the uncinate fasciculus. I converted lh.superiortemporal.label to lh.superiortemporal.nii.gz using mri_label2vol I used probtract as probtrackx -x lh.lateralorbitofrontal.label --mesh=lh.white.asc --xfm=fs2fa_warp --invxfm=fa2fs_warp --seedref=brain.ni.gz -s ../merged -m=../nodif_brain_mask --dir=fibretrack -l -c 0.2 -S 2000--steplength=0.5 -P 5000 --forcedir --opd --waypoints=lh.superiortemporal.nii.gz --stop=non_wm_mask.nii.gz But I got empty tract. I think that I got empty result since of --waypoints option. lh.superiortemporal is a ROI in the coritical region, it is supposed that there is no tract passing this ROI. My understanding is that Probtractx doesn't use an FA threshold and use consistent directions to follow for tractography. Generally, the fibre tracks will turn away from the cortical regions when they are in the vovels near the cortical regions. So the target mask can not be in the cortical regions. But I think that my understanding is wrong, since I knew there were success using the target mask in the cortical regions from old archive. Could you please help me clarify? For my case, what should I do on your experience? Thank you very much Best Shugao On Fri, Nov 12, 2010 at 12:17 PM, Saad Jbabdi <[log in to unmask]> wrote: > mri_label2vol uses the coordinates to know where it is in the brain. > probtrackx uses the coordinates to track, but needs to know the surface > label in order to calculate the normal to the surface... > in your label file, all the coorinate IDs are set to -1. I assumed that was > because you used tkmedit instead of tksurfer (I might be wrong here). > > saad. > > > On 12 Nov 2010, at 16:40, Shugao Xia wrote: > > Hi, Saad, > It does work, since tksurfer doesn't work in my mechine, I can not > check the label in tksurfer. > but using mri_label2vol convert label file into .nii.gz, feed into fslview > with brain.nii.gz, > they look matched. I can't understand why the label file is not valid. > > Thank you for your patience and help > > Best > > Shugao > > On Fri, Nov 12, 2010 at 11:02 AM, Saad Jbabdi <[log in to unmask]> wrote: > >> Hi >> Your label file is not valid: the vertex coordinates do not point at >> locations in lh.white. >> You need to create your label file using tksurfer, not tkmedit. >> Or if you are using the aparc segmentation, you can create a label file >> for each parcel using mri_annotation2label >> >> Cheers, >> Saad >> >> On 12 Nov 2010, at 15:36, Shugao Xia wrote: >> >> I have uploaded my data. The ref # 162936 >> Thank you very much for your help >> >> Best >> >> Shugao >> >> On Thu, Nov 11, 2010 at 5:59 PM, Saad Jbabdi <[log in to unmask]> wrote: >> >>> Can you upload your (tar-zipped) data (including the label files and >>> the lh.white.asc) to www.fmrib.ox.ac.uk/cgi-bin/upload.cgi >>> and send the 6-digit ref number? >>> >>> Saad. >>> >>> >>> On 11 Nov 2010, at 22:20, Shugao Xia wrote: >>> >>> Hi, Saad, >>> >>> I couldn't get the right ROI label from Freesurfer. >>> >>> I did it as following: >>> 1) tkmedit 001 brain.mgz >>> 2) File ---> load segmentation ---> select aparc+aseg.mgz and for color >>> table, choose FreesurferColorLUT.txt >>> 3) click on the label needed (lh_...) >>> 4) Tools ---> segmentation ---> select main segmentation lable at cursor >>> 5) File ---> label---> save lable as... >>> >>> and mris_convert $SUBJECT_DIR/001/surf/lh_white lh_white_asc >>> >>> when I ran the probtrackx with this label >>> >>> the error messaged appeared: >>> load mesh files >>> Segmentation fault >>> >>> I tried several different labels, but still didn't work. >>> >>> what should I get the right label? Thank you very much for your help >>> >>> Best >>> >>> Shugao >>> >>> On Thu, Nov 11, 2010 at 3:34 PM, Saad Jbabdi <[log in to unmask]>wrote: >>> >>>> Hi Shugao, >>>> >>>> On 11 Nov 2010, at 20:29, Shugao Xia wrote: >>>> >>>> Hi, Saad >>>> >>>> I update the fsl version to fsl 4.1.6. But I ran probtrackx as: >>>> >>>> probtrackx -x lh_orbital.label --mesh=lh.white.asc --xfm=fs2fa_warp >>>> --invxfm=fa2fs_warp >>>> --seedref=brain.ni.gz -s ../merged -m=../nodif_brain_mask >>>> --dir=fibretrack -l -c 0.2 -S 2000 >>>> --steplength=0.5 -P 5000 --forcedir --opd --waypoints=lh_suptem.nii.gz >>>> >>>> the error messages still appeared: >>>> load mesh files >>>> Read othererror opening file >>>> >>>> >>>> this error means probtrackx couldn't read you label file (the -x >>>> option). >>>> either the path is wrong, or the file is not in the right format. >>>> >>>> >>>> >>>> What am I doing wrong about the above command? >>>> >>>> I try to track the fibre form the orbitalfrontal to the superior >>>> temporary. ROIs lh_orbital and lh_suptem is generated from freesurfer. >>>> 1) I am not sure which mask option as target mask, --waypoints or >>>> --stop? are they same or what is different? >>>> >>>> >>>> yes they are very different. the doc explains how they are different. >>>> you can use either one or both or none depending on your experiment. >>>> >>>> >>>> 2) can we use this two lables as seed mask in multiple masks? >>>> >>>> >>>> sorry, this is not (yet) available. you can only use a label file in >>>> "single mask" mode. >>>> >>>> >>>> cheers, >>>> saad. >>>> >>>> >>>> Thank you very much >>>> >>>> Shugao >>>> >>>> >>>> >>>> On Wed, Nov 10, 2010 at 10:29 AM, Saad Jbabdi <[log in to unmask]>wrote: >>>> >>>>> Hi >>>>> You cannot use FNIRT warpfields in probtrackx (even from the GUI) if >>>>> you are using an older version. >>>>> Please install a newer version and try again. >>>>> Or use a FLIRT transform matrix for seed->diffusion. >>>>> >>>>> Saad. >>>>> >>>>> On 10 Nov 2010, at 15:23, Shugao Xia wrote: >>>>> >>>>> Thank you very much. >>>>> I tried to run probtrackx on GUI, but there was a error message: >>>>> >>>>> ../bin/probtrack --mode=seedmask -x ../lh_orbital.nii.gz -l -c 0.2 -S >>>>> 2000 >>>>> -steplength=0.5 -P 5000 --xfm=../fs2fa_warp.nii.gz >>>>> --stop=../non_wm_mask.nii.gz >>>>> --forcedir --opd -s ../bedpostX/merged -m ../nodif_brain_mask >>>>> --dir=probtrack >>>>> --targetmasks=../lh_suptem.nii.gz --os2t >>>>> >>>>> Log directory is : ../probtrack >>>>> 1_1 >>>>> 1_2 >>>>> 1_3 >>>>> 2_1 >>>>> 2_2 >>>>> 2_3 >>>>> 7 >>>>> >>>>> Error: terminate called after throwing an instance of >>>>> 'NEWMAT:IndexException' >>>>> >>>>> Do you have any idea what I was doing wrong? Thank you again. >>>>> BTW: lh_orbital.nii.gz lh_suptem.nii.gz non_wm_mask.nii.gz are >>>>> generated from Freesurfer >>>>> >>>>> Shugao >>>>> >>>>> >>>>> On Wed, Nov 10, 2010 at 9:24 AM, Saad Jbabdi <[log in to unmask]>wrote: >>>>> >>>>>> You need a more recent version (>=4.1.5). >>>>>> Note: the doc on the web is only valid for the current release. >>>>>> The doc installed with your FSL version (e.g. >>>>>> $FSLDIR/src/fdt/doc/index.html ) is valid for your version. >>>>>> >>>>>> Cheers, >>>>>> Saad. >>>>>> >>>>>> >>>>>> On 10 Nov 2010, at 14:20, Shugao Xia wrote: >>>>>> >>>>>> I checked and I ran the probtrackx in the directory in which the >>>>>> file lh_orbital.label lies. And if I input option >>>>>> -invxfm, there was a message: >>>>>> --invxfm: option doesn't exist >>>>>> >>>>>> is fsl v4.10 version too old? >>>>>> >>>>>> >>>>>> On Tue, Nov 9, 2010 at 3:38 PM, Saad Jbabdi <[log in to unmask]>wrote: >>>>>> >>>>>>> Looks like it couldn't locate your input in -x. >>>>>>> Make sure you provide the full path to the file lh_orbital.label >>>>>>> (unless it is in the directory you run probtrackx from) >>>>>>> >>>>>>> saad >>>>>>> >>>>>>> On 9 Nov 2010, at 20:08, Shugao Xia wrote: >>>>>>> >>>>>>> Thank you very much. >>>>>>> When I ran probtrackx, >>>>>>> probtrackx -x lh_oribital.label --mesh=lh.white.asc --xfm=fs2fa_warp >>>>>>> --seedref=brain.nii.gz >>>>>>> -s ../bedpostX/merged -m ../bedpostX/nodif_brain_mask >>>>>>> --dir=probtrackx >>>>>>> -l --pd -c 0.2 -S 2000 --steplength=0.5 -P 5000 --forcedir --opd >>>>>>> >>>>>>> some errors happened: >>>>>>> Log directory is : probtrackx >>>>>>> Loading mesh files >>>>>>> Read othererror openning file >>>>>>> >>>>>>> but I ran as >>>>>>> probtrackx -x lh_oribital.nii.gz --xfm=fs2fa_warp >>>>>>> --seedref=brain.nii.gz >>>>>>> -s ../bedpostX/merged -m ../bedpostX/nodif_brain_mask >>>>>>> --dir=probtrackx >>>>>>> -l --pd -c 0.2 -S 2000 --steplength=0.5 -P 5000 --forcedir --opd >>>>>>> >>>>>>> The errors were following : >>>>>>> Log directory is : probtrackx >>>>>>> exit without doing anything >>>>>>> >>>>>>> Could you please tell me where I was doing wrong? There is no option >>>>>>> -invxfm, >>>>>>> I use fsl v4.10 version, is it old such that there is no option >>>>>>> -invxfm in probtrackx? >>>>>>> >>>>>>> Thank you >>>>>>> >>>>>>> Best >>>>>>> >>>>>>> Shugao >>>>>>> On Tue, Nov 9, 2010 at 12:31 PM, Saad Jbabdi <[log in to unmask]>wrote: >>>>>>> >>>>>>>> Hi >>>>>>>> No you need your seedref to be in the "freesurfer space". >>>>>>>> Follow the instructions on the website! >>>>>>>> www.fmrib.ox.ac.uk/fsl/fdt/fdt_surface.html >>>>>>>> >>>>>>>> Saad. >>>>>>>> >>>>>>>> >>>>>>>> >>>>>>>> On 9 Nov 2010, at 17:22, Shugao Xia wrote: >>>>>>>> >>>>>>>> > Dear FSL Experts, >>>>>>>> > I am working on running probtrackx from freesurfer. The label >>>>>>>> file and GM/WM boundary surfuce are generated from freesurfer. >>>>>>>> > when running probtracks: >>>>>>>> > >>>>>>>> > probtrackx -x mylabel.label --mesh=lh.white.asc >>>>>>>> --seedref=brain.nii.gz --xfm=fs2fa_warp.nii.gz --invxfm=fa2fs_warp ... >>>>>>>> > >>>>>>>> > I know there are different directions in Freesurfer and FSL. I am >>>>>>>> not sure whether I need convert barin.nii.gz same as DTI image using >>>>>>>> fslswapdim? If I need, do I need to change mylabel.label or lh.white.asc >>>>>>>> too? >>>>>>>> > >>>>>>>> > Thank you in advance >>>>>>>> > >>>>>>>> > All the best >>>>>>>> > >>>>>>>> > Shugao >>>>>>>> >>>>>>>> -- >>>>>>>> Saad Jbabdi >>>>>>>> University of Oxford, FMRIB Centre >>>>>>>> >>>>>>>> JR Hospital, Headington, OX3 9DU, UK >>>>>>>> (+44)1865-222466 (fax 717) >>>>>>>> www.fmrib.ox.ac.uk/~saad >>>>>>>> >>>>>>> >>>>>>> >>>>>>> -- >>>>>>> Saad Jbabdi >>>>>>> University of Oxford, FMRIB Centre >>>>>>> >>>>>>> JR Hospital, Headington, OX3 9DU, UK >>>>>>> (+44)1865-222466 (fax 717) >>>>>>> www.fmrib.ox.ac.uk/~saad >>>>>>> >>>>>>> >>>>>>> >>>>>>> >>>>>>> >>>>>>> >>>>>>> >>>>>>> >>>>>>> >>>>>>> >>>>>>> >>>>>> >>>>>> -- >>>>>> Saad Jbabdi >>>>>> University of Oxford, FMRIB Centre >>>>>> >>>>>> JR Hospital, Headington, OX3 9DU, UK >>>>>> (+44)1865-222466 (fax 717) >>>>>> www.fmrib.ox.ac.uk/~saad >>>>>> >>>>>> >>>>>> >>>>>> >>>>>> >>>>>> >>>>>> >>>>>> >>>>>> >>>>>> >>>>>> >>>>> >>>>> -- >>>>> Saad Jbabdi >>>>> University of Oxford, FMRIB Centre >>>>> >>>>> JR Hospital, Headington, OX3 9DU, UK >>>>> (+44)1865-222466 (fax 717) >>>>> www.fmrib.ox.ac.uk/~saad >>>>> >>>>> >>>>> >>>>> >>>>> >>>>> >>>>> >>>>> >>>>> >>>>> >>>>> >>>> >>>> -- >>>> Saad Jbabdi >>>> University of Oxford, FMRIB Centre >>>> >>>> JR Hospital, Headington, OX3 9DU, UK >>>> (+44)1865-222466 (fax 717) >>>> www.fmrib.ox.ac.uk/~saad >>>> >>>> >>>> >>>> >>>> >>>> >>>> >>>> >>>> >>>> >>>> >>> >>> -- >>> Saad Jbabdi >>> University of Oxford, FMRIB Centre >>> >>> JR Hospital, Headington, OX3 9DU, UK >>> (+44)1865-222466 (fax 717) >>> www.fmrib.ox.ac.uk/~saad >>> >>> >>> >>> >>> >>> >>> >>> >>> >>> >>> >> >> -- >> Saad Jbabdi >> University of Oxford, FMRIB Centre >> >> JR Hospital, Headington, OX3 9DU, UK >> (+44)1865-222466 (fax 717) >> www.fmrib.ox.ac.uk/~saad >> >> >> >> >> >> >> >> >> >> >> > > -- > Saad Jbabdi > University of Oxford, FMRIB Centre > > JR Hospital, Headington, OX3 9DU, UK > (+44)1865-222466 (fax 717) > www.fmrib.ox.ac.uk/~saad > > > > > > > > > > >