search

danjonpeterson / dti_preproc

preprocessing scripts for diffusion imaging
10 stars 8 forks source link

Incorrect number of words in bvecs file #5

Open atbarr opened 7 years ago

atbarr commented 7 years ago

I found your code after reading a paper on using ANTs to correct for susceptibility artifacts in diffusion images. I'm running into an issue when I run the motion_correct.sh script The script gives an error about the number of values in bvec file. However the text file contains 3 rows with 55 words each.

Could you provide some help into the issue I'm having. I've looked at the source code and as far as I can tell I'm meeting all of the criteria.

bvals_MJ0054.txt bvecs_MJ0054.txt

" ERROR: bvecs file contains 165 words, it should be 3 x 55 = 165 words "

danjonpeterson commented 7 years ago

First of all, thank you for your interest!

Secondly, that certainly is confusing. I downloaded your files and ran them with a shortened diffusion run, and it seemed to work fine for me. Perhaps there is some difference between the way your computer is interpreting the conditional and mine. What shell are you running?

Alternately, since we have determined that your bvecs contains the correct number of lines, you COULD just comment out the offending line (#146) and see how that works. But I'd rather actually fix the problem if we can.

bassamalfatly commented 3 years ago

Hi Daniel, I also have a similar problem with the motion_correct.sh showing an error similar to the one above! any clue on that? I am using MacOS bash, thanks in advance. Bassam

danjonpeterson commented 3 years ago

HI Bassam - that's annoying! If there's some way you can send me the files, I'd be glad to take a look. This was developed on CentOS, so there may be some issue there - It could have something to do with strange whitespace. Again, you could also comment out the line, but that's unsatisfying. If lots of people are having this problem I'd rather fix it.

Thanks for your interest, -Dan

atbarr commented 3 years ago

These messages are going to the wrong person.

On Feb 17, 2021, at 5:10 PM, Daniel Jon Peterson notifications@github.com wrote:

 HI Bassam - that's annoying! If there's some way you can send me the files, I'd be glad to take a look. This was developed on CentOS, so there may be some issue there - It could have something to do with strange whitespace. Again, you could also comment out the line, but that's unsatisfying. If lots of people are having this problem I'd rather fix it.

Thanks for your interest, -Dan

— You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub, or unsubscribe.

bassamalfatly commented 3 years ago

Hi Daniel, thanks for you reply, these are the bvals, bvecs, they are actually from a subject from the PPMI dataset, I have downloaded the DICOM images and imported them to nii, bvals bvecs using https://github.com/rordenlab/dcm2niix

dti_bvals.txt dti_bvecs.txt