[divarada]

Can you try adding masks for structural and diffusion data and try running the BDP command.

There are two flags I want you to add and let me know if the error still persists:
1. –dwi-mask <filename.nii.gz> : Specifies the filename of the brain-mask file for diffusion data.
2. –t1-mask <filename.nii.gz> : Specifies the filename of the brain-mask file for input T1-weighted image.This mask can be same as the brain mask generated during BrainSuite extraction sequence.

More details of flags can be found at http://brainsuite.org/processing/diffusion/flags/

[divarada]

BDPsummary should have been generated in the same folder as the bfc file but since in your case it did not generate then most likely BDP exited before that step.

I repeated your command with a dataset I had :
bdp.exe 2523412\2523412.bfc.nii.gz –tensor –fract –frt –nii 2523412\2523412.DTI64.nii -b 2523412\2523412.DTI64.bval -g 2523412\2523412.DTI64.bvec

and it was able to find my file and went beyond the step that is crashing for you. So looks to be a syntax issue.

Even though I am not sure, here are quick pointers that might help you:
– check if all files exist and are accessible to bdp
– check space, double dashes
– remove fract and frt, just run tensor
– try a different input dataset

[divarada]

This looks like matlab runtime has crashed – have you tried running BDP on a different machine or reinstalling matlab runtime ?

[divarada]

Do your tensors look okay ? You can visualize it on BrainSuite GUI by uploading the name.eig,nii.z file.

It will be beneficial to also try the same steps with a different dataset and see if FA looks okay or noisy

[divarada]

There must be a BDPSummary.txt generated. It will have details of the command that was used such as this one:

COMMAND USED:

bdp.exe 2523412.bfc.nii.gz
–nii 2523412.DTI64.nii.gz
–bval 2523412.DTI64.bval
–bvec 2523412.DTI64.bvec
–tensors
–frt
–fract

Approximate processing time: 19.82 minutes

Can you please copy the text from there ?

[divarada]

Hi, currently we do not support motion correction. One way you can correct for motion is by rigid registration of all diffusion weighted images with one reference image (like a b=0 image). You can use our affine registration addon tool to achieve this – https://neuroimage.usc.edu/neuro/Resources/BDPAddons#affineReg

[divarada]

BDP does not support saving the diffusion data in T1 space mainly due to memory requirements and since the corresponding data is not saved it does not save the rotated bvecs/bmats either.

However, we do have additional tools to perform the rigid registration.

Here is the link to the rigid registration source code: http://neuroimage.usc.edu/~chitresh/BDPinternal/BDP_addons/register_files_affine_v12.zip

You can run this to perform rigid registration and get diffusion files in T1 space.

However, to be able to obtain the rotated bvecs, you will need to make changes to BDPs code and rebuild it. Please refer to estimate_tensors_mprage.m file in BDP’s source code. There you can see how BDP internally rotates the bmatrices. Basically, you will have to save the rotated bvecs/bmats there and rebuild BDP or replicate that code in your current setup.

[divarada]

Since bdp was killed, it is not crashing internally.
One possible reason might be that the system has run out of memory.

[divarada]

Can you share the dwi mask file (dropbox ?)? This is most likely a corner case related to the masking file specified and debugging this issue is specific to the particular images that you are using.

[divarada]

Can you try without giving the dwi-mask and/or t1-mask then ?

Generally we have seen this when the mask is creating some issues. I am not sure if its the same reason in your case, so we will have to try a few things.

[divarada]

typo : –dwi-mask (there are two hyphens)

[divarada]

This issue generally happens when the masking code fails in BDP – since this is a heuristic approach it is possible that it fails for certain cases. Can you try specifying the brain mask file diffusion data ? Use –dwi-mask flag for specifying the mask file.

[divarada]

You can use the following function available in our source code release to generate eig.nii.gz files that can be visualized on BrainSuite.

function generate_eig_file(L1, L2, L3, V1, V2, V3, outputFileName)
%GENERATE_EIG Generate .eig file. Eigenvalues (L1, L2, L3) &
% Eigenvectors(V1, V2, V3) can be nii structures or .nii.gz or .nii files.
%
% outputFileName – file name with/without extension (with full path)
%

As you can see in the description, it just needs the tensor eigenvalues and eigenvectors. The function concatenates these files in a certain way that BrainSuite understands. It should be clear when you just look at this function.

Thanks.

[divarada]

Yes, BDP reorients the volume to RAS format. While performing these orientations BDP applies the appropriate rotations to the bvecs and bMatrices tables.

[divarada]

Sorry for the delay.

BDP refines the DWI mask that you provide to remove CSF region. Your logs and the fact that you have *.rearranged.b0s.nii result saved indicate that this step is crashing for the mask that you provided.
If you can share the mask we can run it here and tell you what went wrong. Masking methods used in BDP are heuristic approaches that can fail for certain cases.

Also, the following information will be helpful :
Did you use the same mask before when it ran successfully for you ?
Can you try running BDP without the DWI mask (no –dwi-mask) – Does it still throw an error ?

[Author]

Posts