search

BrainModes / TVB-empirical-data-pipeline

Empirical processing pipeline for The Virtual Brain
http://dx.doi.org/10.1016/j.neuroimage.2015.03.055
22 stars 19 forks source link

Running pipeline with dwMRI of 30 directions #6

Open azin-d opened 7 years ago

azin-d commented 7 years ago

Hi there,

Is there anyway to run the pipeline with DTI of only 30 directions. As we tried it the process failed with error message "ERROR! Not enough distinct directions in your dwMRI data! Sorry." There was no results folder and couldn't find SC or FC matrices.

Thank you, Azin

azin-d commented 7 years ago

Hi again,

I tried running the pipeline with DTI data of 64 directions but it still gives us the error message that "ERROR! Not enough distinct directions in your dwMRI data! Sorry." Is there anyway we can debug this?

Thank you, Azin

srothmei commented 7 years ago

Hi Azin,

sorry for the late response.

How many different gradient directions where applied during your dwMRI scanning? You need a certain amount of gradient-directions in order to perform the spherical deconvolution before tractography. Note that those directions i.e. the vectors that depict it must can be redundant or securing multiple times.

Could you post your bvec-file please? It should be somewhere in your subject folder. Thanks

azin-d commented 7 years ago

Hi Simon,

Thank you for helping us. Sorry for my late reply. We used 62 directions. I noticed that both bvec-file and bval files are empty but attached you can find the log files.

Thank you, Azin

On Sat, Jun 24, 2017 at 12:36 AM, Simon Rothmeier notifications@github.com wrote:

Hi Azin,

sorry for the late response.

How many different gradient directions where applied during your dwMRI scanning? You need a certain amount of gradient-directions in order to perform the spherical deconvolution before tractography. Note that those directions i.e. the vectors that depict it must can be redundant or securing multiple times.

Could you post your bvec-file please? It should be somewhere in your subject folder. Thanks

— You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub https://github.com/BrainModes/TVB-empirical-data-pipeline/issues/6#issuecomment-310813517, or mute the thread https://github.com/notifications/unsubscribe-auth/AEzUQPKF_O1Bu2EeKDVo6vjYByxhjV5pks5sHJJEgaJpZM4N1rGP .

lauramediavilla commented 5 years ago

Hi Simon, I have got possibly the same problem as Azin. We used 61 directions and neither my bvec nor my bval files are empty (see them attached), however I got the same error message (ERROR! Not enough distinct directions in your dwMRI data! Sorry).

Is there any way to solve this? Thank you in advance.

Laura

bvec.txt bval.txt

srothmei commented 5 years ago

Hi,

i checked your file (bvec) and it actually seems perfectly okay. I also ran it through the code section in preprocDK.sh (lines 164 to 182), giving me no error whatsoever. The resulting lmax parameter for MRtrix is 8 in this case.

Could you please add the following line to the preprocDK.sh script, just directly after the command distinctDirections="$(( $(sort bvec | uniq | wc -l) - 1 ))" on line 164: echo "Dist dir: $distinctDirections"

If you rerun the script now, it should display your distinct connection count right before it exits with the aforementioned error.

Furthermore, your bval file seems a bit strange since the values are slightly drifting around the "usual" values one would expect, like 0,1000,1500,2000 etc.

Best Simon

lauramediavilla commented 5 years ago

Hi Simon,

Thank you for you nice reply! The error came from running the pipeline using the NSG portal, so I am afraid I cannot change that code. Maybe that is an issue to solve with the NSG developers, because until now I am unable to tun the pipeline using NSG.

I am also running the pipeline in my workstation, and I have not had that error yet (I am having others). However, the preprocDK script I have downloaded (from this Github, TVB-empirical-data-pipeline-OAR) does not contain the line you mentioned: distinctDirections="$(( $(sort bvec | uniq | wc -l) - 1 ))"

The bvals values come from a b value of 1000 s/mm2 that was set on the scanner. I am not sure why they drift (I asked my supervisor).

Best, Laura

srothmei commented 5 years ago

Hi Laura,

sorry for the delayed response.

However, the preprocDK script I have downloaded (from this Github, TVB-empirical-data-pipeline-OAR) does not contain the line you mentioned:

Yeah the OAR-Branch is a bit behind considering code-commits compared to the NSG-Branch. Did you try the NSG branch on your local machine?