Print

Print
Well I looked at the log and the command is almost the same. The difference is that through the command line I first ran fslmaths to add all the rois and the resulting image i introduced it in the probtrackx command after the -x, and took out the --network as you said. 

When it is done through the GUI, it adds the --network and after the -x it puts the path of a txt file that contains the list of rois. I guess that is the reason for getting different answers but then, which one is the correct sentence? which one should I use? Because the results are not just slightly different, they are very different. I do use the same computer.

This is what I wrote when I ran from the command line:
fslmaths -dt double "path_roi1" -add ""path_roi2" "result"

probtrackx --mode=seedmask -x "result" -l -c 0.2 -S 2000 --steplength=0.5 -P 5000 --forcedir --opd -s "bedpostdir/"merged -m "bedpostdir/"nodif_brain_mask.nii  --dir="outputdir" --omatrix1

And this is what the log shows when I run it from the GUI:
probtrackx --network --mode=seedmask -x /.../masks.txt -l -c 0.2 -S 2000 --steplength=0.5 -P 5000 --forcedir --opd -s "bedpostdir/"merged -m "bedpostdir/"nodif_brain_mask --dir="outputdir" 

On Tue, Mar 29, 2011 at 12:21 PM, Saad Jbabdi <[log in to unmask]> wrote:
You mean you ran it twice in the GUI or once in the GUI and once in the command line? If it is the latter, then I suspect you are not using quite the same options (e.g. you may have forgotten the --loopcheck, which is default in the GUI). Look at probtrackx.log to make sure.
Also, you may get (slightly) different answers if you run probtrackx on different machines, simply because of differences in random number generators.

Saad.


On 29 Mar 2011, at 09:57, Monica Giraldo wrote:

Hello, I've just ran the probtrackx through the GUI with the same parameters (the default ones), the same ROIs, and it gives different answer. now the waytotal has these two numbers: 80075 and 138465. Is this normal? which results should I trust?

On Mon, Mar 28, 2011 at 5:03 PM, Jose Angel Pineda Pardo <[log in to unmask]> wrote:
Try to use that

"gunzip path_to_nii.gz"

Jose


Mon, Mar 28, 2011 at 4:58 PM, Monica Giraldo <[log in to unmask]> wrote:
Sorry I have one last quesiton, is it possible to uncompress an image with FSL through the command line? Because I changed a little bit the script for generating the matrix and I need to introduce the images in .nii, because I got a problem when I tried to use .nii.gz

Thanks,


On Mon, Mar 28, 2011 at 4:29 PM, Jose Angel Pineda Pardo <[log in to unmask]> wrote:
Hi Monica,

it seems you are just using two seed regions. So in that case it will be correct.

Cheers

Jose


On Mon, Mar 28, 2011 at 4:10 PM, Monica Giraldo <[log in to unmask]> wrote:
Hi Saad, thanks for your help. Well, my seeds are already in the DTI space. I did all the pre processing of the images with SPM before running anything in FSL and all the images used are in the same space as the DTI. That's why im not using the --xfm option in probtrackx.

I tried anyway running your script and I did got a result but I don't know it it's correct. I got the following matrix:
ConMat = 
62470809    321160
461766        46066148

but why the waytotal file shows the number 1.5015e+07? If it is not the sum of the numbers of the matrix?

It may be something else wrong?

Thankx,

On Mon, Mar 28, 2011 at 2:15 PM, Saad Jbabdi <[log in to unmask]> wrote:



sorry, I forgot another thing. and MATLAB doesn't recognize the command read_avw, is there any chance that you wrote it wrong? Do I have to install somehting in matlab to run that command?

that is why I've added the line "addpath([getenv('FSLDIR') '/etc/matlab']);" in the script that I suggested to you. This tells matlab to look for the appropriate directory where read_avw is coded.


Saad.



On Mon, Mar 28, 2011 at 12:46 PM, Monica Giraldo <[log in to unmask]> wrote:
Hello Saad, I just tryed doing what you told Jose. I executed first the following command for adding all the seed masks: 

fslmaths -dt double "mask1_path" -add "mask2_path" "resultImage_path"

I put the type of the images as double because when I open them in MATLAB it says that they are double type.

Then, I executed the probtrackx command with the following options:

probtrackx --mode=seedmask -x "resultImage_path" -l -c 0.2 -S 2000 --steplength=0.5 -P 5000 --forcedir --opd -s "bedpostDirectory+/merged" -m "nodif_brain_mask_path"  --dir="output directory" --omatrix1

Well, it did created the files it should have created: the coords_for_fdt_matrix1, fdt_matrix1, fdt_paths and waytotal. But there is something strange, although it does find tracks because in the waytotal file it appears the number 1.5015e+07 when I open the fdt_paths there is nothing! The other images dimensions are correct. And when I open the images I can visualize something.

Other quesiton I have is, in your code when you write seeds = {'seed1','seed2',...}; 'seed1' and 'seed2' are the paths of the seed images or what?

Thanks,


--
Monica Giraldo
Master student in Biomedical Engineering
Universidad de Barcelona - Universidad Politécnica de Cataluña




--
Monica Giraldo
Master student in Biomedical Engineering
Universidad de Barcelona - Universidad Politécnica de Cataluña


--
Saad Jbabdi
University of Oxford, FMRIB Centre

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













--
Monica Giraldo
Master student in Biomedical Engineering
Universidad de Barcelona - Universidad Politécnica de Cataluña





--
Monica Giraldo
Master student in Biomedical Engineering
Universidad de Barcelona - Universidad Politécnica de Cataluña





--
Monica Giraldo
Master student in Biomedical Engineering
Universidad de Barcelona - Universidad Politécnica de Cataluña


--
Saad Jbabdi
University of Oxford, FMRIB Centre

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













--
Monica Giraldo
Master student in Biomedical Engineering
Universidad de Barcelona - Universidad Politécnica de Cataluña