Dear All,

I have to deal with Siemens syngo.via DTI DICOMs for a recent project and, thus, had to switch to a newer version of dcm2niix. The old version seems not to be able to extract the diffusion gradients (bvecs/bvals) from these new files.

I now experience a "segmenation fault" error in a Centos 7 enviroment when I do a conversion. The help text is displayed fine, so the programm does run, but any conversion immediately gives this crash.

I could reproduce this with the dcm2niix version available via mricrogl (2018-08-03) and the 2018-11-25 from the dcm2niix GitHub (including the LinuxCompat version). This does NOT happen with the version I was using previously (2017-05-28, also from mricrogl).

I know that Centos is a bit special with old versions used for some components. Still, I believe RHEL/SL/Centos/... 7 is not totally outdated. The "ldd"'s seem all to be there.

Any suggestions how to get this running?

Many thanks,

Niels

Hello,

I use dcm2niix on Centos 7 and am not having that problem. I just ran a test suite on a wide variety of MR scans with dcm2niix 20190116 and it passed.

Do these Siemens scans still have their CSA headers (*large* structures in 0029, 0010 or 0029, 1010)?

Rob

1.) Can you test your dataset with the development branch: (1.0.20190323)

git clone -b development https://github.com/rordenlab/dcm2niix.gi...
cd dcm2niix
mkdir build && cd build
cmake ..
make
cd bin
./dcm2niix
Chris Rorden's dcm2niiX version v1.0.20190323 Clang10.0.0 (64-bit MacOS)


2.) If this still fails, can you create a new issue on Github
  https://github.com/rordenlab/dcm2niix/issues
A sample image demonstrating the issue would help. You could send a sample set directly to me via box or DropBox.

By the way, what Siemens hardware was this from. I know that Tuebingen has a Siemens Vida, and this generates enhanced DICOMs (e.g. software XA10) while all other Siemens equipment generate conventional DICOMs (e.g. software VB13, VB17, VE11, etc.). If this is XA10A data, be aware that Siemens does not include a CSA header (so you will not get phase encoding polarity, readout time, etc). Further, if XA10 data is saved as "mosaic" or "de-identified" formats, the DICOMs will omit fields for determining gradient direction, patient position, etc.

Thanks Chris for the pointers,

to sum it up for the forum:

The segmentation fault was solved by compiling/making the recent developmental version on our Centos7 system. The seg fault did happen on all scans, not just the "enhanced" DICOMs we tried to convert in the first place.

Furthermore, also these DICOMs were not really standard (coming from a mosaic/CSA header scanner) and going through syngio.via (partially converting them to the "enhanced" standard). The develeopmental version did create bvec/bval files eventually. However, we had to disable the -s flag. It seems that bval/bvec are not generated if slices are missing.

So, after all, it does work to process even these "difficult" files.