 |
 |
Hi It is probably still tracking. If you use --verbose=1 you should be able to see the seed vertices being examined one by one. The only thing you can do to speed it up is reduce the number of samples (that should give you an idea of whether it has worked, and how long it should take with the full number of samples). I am afraid you cannot change the seed ref (but that is not the thing that is causing probtrackx2 to be slow). Cheers Saad On 26 Oct 2012, at 05:58, Alistair Perry wrote: > Ok thankyou, I removed the --omatrix3 option and it has gone past the > stage I was stuck at initally. > > However, for the last few hours it has been stuck as this point, > straight after loading the bedpostx samples. > > Done loading samples. > Surface seeds > surface 0 > > Should this happen? I know it is a memory issue, but i have recently > upgraded to a 64GB RAM machine. I know you can speed up the process by > reducing the number of vertices, but can I also use a smaller seed ref > image (i.e rawavg.nii instead of orig.nii) to speed it up. Obviously > the rawavg.nii file is in structural instead of freesurfer space, but > does this matter? > > Thanks, > > Alistair > > > On Thu, Oct 25, 2012 at 7:15 PM, Saad Jbabdi <[log in to unmask]> wrote: >> Hi - Are you sure you intended to use --omatrix3? This option requires you to set a --target3 file. >> >> Cheers >> Saad >> >> On 25 Oct 2012, at 05:34, Alistair Perry wrote: >> >>> Sorry, still having issues. >>> >>> To rehash what I am wanting to achieve, I am trying to perform >>> whole-brain tractography analysis. >>> >>> I have preprocessed DTI and T1 scan using combination of >>> freesurfer/fsl, so I have sucessfully created bedpostx and all >>> recon-all files. >>> >>> I have obtained 68 cortical label labels from performing >>> mri_annotation2label, and then label2surf. Obviously these are all in >>> conformed space. >>> I have used the instructions as per your wiki page in obtaining >>> freesurfer<>structural<>dti(S0 image) transformation files >>> >>> I have all the working parts and here is the command: N.B all of the >>> working files have been placed in the same directory >>> >>> probtrackx2 -s ./merged -m ./nodif_brain_mask.nii -x ./rois.ascii --seedref=./or >>> ig.nii --xfm=./fs2dti.mat --targetmasks=./rois.ascii --dir=../probtrackresults/ >>> --forcedir --s2tastext --omatrix1 --omatrix3 >>> >>> This is the following error message I receive after initally going >>> through loading all of the seeds: >>> >>> Load bedpostx samples >>> 1_1 >>> 1_2 >>> 1_3 >>> 2_1 >>> 2_2 >>> 2_3 >>> >>> nfibres : 2 >>> nsamples : 50 >>> >>> Done loading samples. >>> does not exist >>> >>> It is interesting that another time (where the command would of been >>> identical, although I can't exactly remember) probtrackx2 went past >>> this stage and halted at a specific label file and exited with the >>> error message again "does not exist" >>> >>> Is there anything wrong you can see with my working parts, more >>> specifically, my label files? I have eliminated errors occuring due to >>> different version of fsl used in pre-processing, or the program not >>> finding the files in the directory folder. >>> >>> >>> >>> >>> On Thu, Sep 20, 2012 at 7:28 PM, Saad Jbabdi <[log in to unmask]> wrote: >>>> Looks like you're running out of RAM. >>>> >>>> On 20 Sep 2012, at 07:56, Alistair Perry wrote: >>>> >>>>> Thanks, still having issues... >>>>> >>>>> I have sucessfully initiated probtrackx2, with the same command as above. >>>>> >>>>> However, after a minute or so after calling at the command line, it >>>>> abruptly terminates. Here is a log: >>>>> >>>>> "Running in seedmask mode >>>>> load seeds >>>>> read gifti >>>>> read scalar data that is float32 >>>>> nz=116523 >>>>> nnz=1601 >>>>> read gifti >>>>> read scalar data that is float32 >>>>> nz=117080 >>>>> nnz=1044 >>>>> read gifti >>>>> read scalar data that is float32 >>>>> nz=115083 >>>>> nnz=3041 >>>>> read gifti >>>>> read scalar data that is float32 >>>>> nz=115740 >>>>> nnz=2384 >>>>> read gifti >>>>> read scalar data that is float32 >>>>> nz=117844 >>>>> nnz=280 >>>>> read gifti >>>>> read scalar data that is float32 >>>>> nz=117842 >>>>> nnz=282 >>>>> read gifti >>>>> read scalar data that is float32 >>>>> nz=114311 >>>>> nnz=3813 >>>>> read gifti >>>>> read scalar data that is float32 >>>>> nz=112058 >>>>> nnz=6066 >>>>> read gifti >>>>> read scalar data that is float32 >>>>> nz=113721 >>>>> nnz=4403 >>>>> Killed"" >>>>> >>>>> Do you think this is an issue with the data quality of the surface >>>>> labels. If so, is there anyway in which I can check? >>>>> >>>>> In addition, I thought that I could get around this issue by >>>>> converting the initial label files (before running label2surf) by >>>>> converting them individually to asc files. However, I am unable to do >>>>> so. >>>>> >>>>> I ran the command: # mris_convert --label lh.bankssts.label >>>>> lh.bankssts lh.white lh.bankssts.asc >>>>> ERROR: unknown file annot file type specified for output: lh.bankssts.asc >>>>> >>>>> It is not specified in the documentation how to do so. >>>>> >>>>> Thanks, >>>>> >>>>> Alistair >>>>> >>>>> >>>>> On Sat, Sep 15, 2012 at 8:32 PM, Saad Jbabdi <[log in to unmask]> wrote: >>>>>> Hi >>>>>> You need an "=" sign whenever an option has two "-" instead of one. >>>>>> >>>>>> For example: " -m mask" but " --xfm=transform.mat" >>>>>> >>>>>> >>>>>> >>>>>> Saad >>>>>> >>>>>> On 14 Sep 2012, at 23:35, Alistair Perry wrote: >>>>>> >>>>>>> Thanks, however I am having issues actually running the tractography. >>>>>>> >>>>>>> This is my command line: >>>>>>> >>>>>>> -s ./bedpostx/2393_20090210.bedpostX/ -m >>>>>>> ./bedpostx/2393_20090210/nodif_brain_mask.nii.gz -x >>>>>>> ./subjects/2393_20090210/convertedsurfaces/rois.ascii --xfm >>>>>>> ./subjects/fs2dti.mat --meshspace freesurfer --targetmasks >>>>>>> ./subjects/2393_20090210/convertedsurfaces/rois.ascii --dir >>>>>>> ./subjects/2393_20090210/probtrackresults/ --forcedir --os2t >>>>>>> --s2tastext >>>>>>> >>>>>>> I get the following error: >>>>>>> >>>>>>> --xfm: Missing non-optional argument! >>>>>>> >>>>>>> I tried to rearrange the command line: >>>>>>> >>>>>>> probtrackx2 -s ./bedpostx/2393_20090210.bedpostX/ -m >>>>>>> ./bedpostx/2393_20090210/nodif_brain_mask.nii.gz -x >>>>>>> ./subjects/2393_20090210/convertedsurfaces/rois.ascii --seedref >>>>>>> ./subjects/2393_20090210/mri/T1.nii --xfm ./subjects/fs2dti.mat >>>>>>> --meshspace freesurfer --targetmasks >>>>>>> ./subjects/2393_20090210/convertedsurfaces/rois.ascii --dir >>>>>>> ./subjects/2393_20090210/probtrackresults/ --forcedir --os2t >>>>>>> --s2tastext >>>>>>> >>>>>>> and get another error: >>>>>>> >>>>>>> --seedref: Missing non-optional argument! >>>>>>> >>>>>>> >>>>>>> Do you have any idea what is going on here? I have looked through the >>>>>>> forums and it is pretty much carbon copy to what other people have >>>>>>> been doing. It doesnt matter whether I rearrange the command line it >>>>>>> will still get stuck on either --xfm or --seedref. Is there another >>>>>>> option I must include if i choose to use either --xfm or --seedref? >>>>>>> >>>>>>> When I >>>>>>> On Thu, Sep 13, 2012 at 9:15 PM, Saad Jbabdi <[log in to unmask]> wrote: >>>>>>>> Hi >>>>>>>> You need to transform your label files to surface files as explained here: http://fsl.fmrib.ox.ac.uk/fsl/fslwiki/FDT/UserGuide#Label_files >>>>>>>> >>>>>>>> Cheers >>>>>>>> Saad >>>>>>>> >>>>>>>> >>>>>>>> On 13 Sep 2012, at 00:26, Alistair Perry wrote: >>>>>>>> >>>>>>>>> I have all the working parts: >>>>>>>>> >>>>>>>>> - DTI image registered to freesurfer image using bbregister and flirt. >>>>>>>>> From here I have created a transformation and inverse matrix >>>>>>>>> - I have used mri_annotation2label to generate a label for each >>>>>>>>> cortical ROI of the aparc parcellation (so I have 68).. >>>>>>>>> >>>>>>>>> Now, the part I do not understand is the ascii text file list of the >>>>>>>>> volumes and/or surfaces. Is this merely just a list of the labels? >>>>>>>>> >>>>>>>>> For example: >>>>>>>>> >>>>>>>>> lh.bankssts.label >>>>>>>>> lh.caudalanteriorcingulate.label >>>>>>>>> >>>>>>>>> Thanks, >>>>>>>>> >>>>>>>>> Alistair >>>>>>>>> >>>>>>>>> On Fri, Sep 7, 2012 at 11:59 PM, Saad Jbabdi <[log in to unmask]> wrote: >>>>>>>>>> Hi Alistair >>>>>>>>>> >>>>>>>>>> This should now be explained better in the new FSL5 doc (released yesterday). Please have a look at the freesurfer section and let me know if you have issues. >>>>>>>>>> >>>>>>>>>> saad >>>>>>>>>> >>>>>>>>>> On 5 Sep 2012, at 04:23, Alistair Perry wrote: >>>>>>>>>> >>>>>>>>>>> Hello FSL experts, >>>>>>>>>>> >>>>>>>>>>> I know this has been posted elsewhere with the original probtrackx. >>>>>>>>>>> >>>>>>>>>>> I am basically trying to create a surface file from the aseg+aparc.mgz cortical parcellation created when running recon-all. I wish to use this surface file as the seed reference file when performing probtrackx2. >>>>>>>>>>> >>>>>>>>>>> I have succeeded sucessfully in transforming the cortical parcellation into diffusion space, and have created a label file for it. >>>>>>>>>>> >>>>>>>>>>> However, I am having difficulty understanding converting the label file into ascii format, as required for probtrackx2. I am aware that you can convert a label file into ascii format using mri_convert, but you have to specify the lh.white surface file created when performing recon-all. This does not make sense to me for one this label file is not in diffusion space. Plus, why would I only specify one surface hemisphere when I have a label file that lists vertices across both hemispheres? >>>>>>>>>>> >>>>>>>>>>> In the end I am asking whether it is possible/and the best method in creating an ascii surface file of the whole cortical parcellation in diffusion space? >>>>>>>>>>> >>>>>>>>>>> Regards, >>>>>>>>>>> >>>>>>>>>>> Alistair >>>>>>>>>>> >>>>>>>>>> >>>>>>>>>> -- >>>>>>>>>> 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